CONFERENCE ABSTRACTS

Infection, Genetics and Evolution 9 (2009) 368–386

Contents lists available at ScienceDirect

Infection, Genetics and Evolution journal homepage: www.elsevier.com/locate/meegid

CONFERENCE ABSTRACTS Keynote Speakers Translation of Genomics in endemic countries Winston Hide SANBI, University Western Cape, South Africa Email: [email protected] In 1994, TDR helped establish five international parasite genome networks and opened the door for scientists from disease endemic countries to participate and collaborate in genome and post-genome projects. A great deal of parasite, vector, and host genome data has been produced. The challenge now is developing an understanding to allow translation of genome sequence in order to deliver novel drug targets, structural predictions, characterisation of biodiversity, reconstruction of metabolic pathways, systems biology and identification of vaccine candidate molecules. In order to take these areas beyond the concept of buzzword, a great deal of knowledge transfer has to occur. This presentation will provide context to bioinformatics research in Africa, outline the challenges inherent in understanding of genome data, explore strategies, and provide examples of ‘translation’. Genomics and Emerging Viral Infectious Diseases: Influenza And Coronaviruses Appolinaire Djikeng Viral Genomics Group, The Institute for Genomic Research (TIGR), A Division of the J. Craig Venter Institute (JCVI), 9704 Medical Center Drive, Rockville, MD 20850, USA Email: [email protected] The majority of emerging diseases threats is of zoonotic origin and of these pathogens, the overwhelming majority are RNA viruses. Examples of important viral threats include agents such as Influenza, SARS, Ebola, Dengue, and Hantavirus. Genomic approaches provide a window to viral evolution and facilitate the study of viral ecology and the relationship between humanity and zoonoses. Due to the relative simplicity of RNA viral genomes the compilation of viral sequence data allows mass comparative genomic analysis and association with epidemiological data at a scale which is currently impossible for other organisms. We are specifically interested in Influenza and Coronoviruses that have the ability to cause diseases 1567-1348/$ – see front matter doi:10.1016/j.meegid.2008.09.003

that significantly affect both humans and animals. In addition we use novel and high throughput genome sequencing and analysis of viral samples in the context of emerging viral infectious diseases. Our technology platform and results will be presented and discussed. The Genome Project of Taenia solium Enrique Morett5*, A. Garciarubio3, R. J. Bobes1, J. C. Carrero1, M. A. Cevallos2, G. Fragoso1, P. Gayta´n3, V. M. Gonza´lez2, M. V. Jose´1, L. Jime´nez4, A. Landa1, C. Larralde1, J. Morales-Montor1, E. Sciutto1, X. Sobero´n3, P. de la Torre1, V. Valde´s5, J. Ya´nez3, J.P. Laclette1y 1 Instituto de Investigaciones Biome´dicas; 2Centro de Ciencias Geno´micas; 3Instituto de Biotecnologı´a; 4Facultad de Medicina; 5 Facultad de Ciencias, Universidad Nacional Auto´noma de Me´xico * Corresponding author. Email: [email protected] y Project coordinator. We have constituted a consortium of key laboratories at the National Autonomous University of Mexico to carry out a full genomic project for Taenia solium. This project is providing invaluable resources for the study of taeniasis/cysticercosis, and, in conjunction with the Echinococcus granulosus and E. multilocularis genome projects of expressed sequence tags (EST’s) is marking the advent of cestode parasites genomics. The T. solium genome size was estimated at about 270 Mb by cytofluorometry on isolated cyton nuclei. In contrast, several probabilistic calculations based on shotgun sequenced genomic clones, resulted in size estimates of 120–140 Mb. The sequencing is being carried out through a combined strategy of 454 pyrosequencing and traditional capillary sequencing. We consider a 20 coverage by pyrosequencing and at least 3 by capillary sequencing from small insert shotgun and fosmid end libraries, to obtain a 95% completeness of the genome. About 40,000 ESTs have been obtained; 14,113 adult and 9157 larva distinct ESTs were made public through GenBank. Unique genes were identified by clustering all EST-fragments with an assembler (minimus). Out of the initial 23,290 ESTs, 19,067 were incorporated into 2564 genes (contigs). We have 6800 ‘‘genes’’, including the contigs, plus 2592 larva and 1611 adult ESTs that remained as solitons. We have found highly expressed genes in both adult and larvae ESTs. There are 349 ‘‘genes’’ with 10 or more sequences that account for 50% of all transcripts. Many are differentially expressed. Approximately 1/3 (2038) of the 6800 genes have a significant match in SwissProt. Gene Ontology (GO) and Gene Ontology broad categories (GoSlim) can be assigned to

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

36% (2448 genes) of our genes. About 27% of the genes have no match in SProt + TreMBL, and could constitute new genes. The highest expressed genes support a very active fat and carbohydrate metabolism. As for the genomic sequences, we are analysing correlated polymorphism: sequences that share non-consensus nucleotides at different positions. These can hint to heterozygosis, gene duplications, and alternative splicing. Our results suggest that the genome of T. solium is not highly repetitive (<7%). One small (53 bp) tandem-repeat represented 0.5% of the genome. Different tetranucleotide repeats of the form (TXXX) account for about 4.5% of the genome. This project is supported by a special grant IMPULSA-UNAM. Data Mining Parasite Genomes Matthew Berriman The Wellcome Trust Sanger Institute, Genome Campus, Hinxton CB10 1SA, UK

369

matics systems applied to the genetic analysis of HIV-1 and other RNA viruses important for human health perspective. 1. BioAfrica website, http://www.bioafrica.net 2. de Oliveira T, Miller R, Tarin M, Cassol S. An integrated genetic data environment (GDE)-based LINUX interface for analysis of HIV-1 and other microbial sequences. Bioinformatics 2003;19(1): 153–4. 3. de Oliveira T, Deforche K, Cassol S, Salminen M, Camacho R, Pavareskevis D, Seebregts C, Snoeck J, VanDamme AM. An automated genotyping system for analysis of HIV-1 and other microbial sequences. Bioinformatics 2005;21:3797–3800. 4. Doherty RS, de Oliveira T, Seebregts C, Danaviah S, Gordon M, Cassol S. BioAfrica’s HIV-1 Proteomics Resource: Combining protein data with bioinformatics tools. Retrovirology 9;2(1):18, 2005. 5. HIV Los Alamos HIV Sequence database, http://www.hiv.lanl.gov 6. Stanford HIV Drug Resistance database, http://hivdb.stanford.edu.

Email: [email protected] 7. RNA Virus database, http://virus.zoo.ox.ac.uk/rnavirusdb/ The BioAfrica Website, Resources and Databases for HIV Tulio De Oliveira1*, Sharon Cassol2 1 MRC Bioinformatics Capacity Development Research Unit, South African National Bioinformatics Institute, The University of the Western Cape, South Africa; 2MRC Inflammation and Immunity Unit, Department of Immunology, University of Pretoria, South Africa * Corresponding author. Email: [email protected] In the age of digital information it is difficult and frustrating to many researchers to identify and use the large number of bioinformatics resources and references that are available to manipulate genetic and other biomolecular data. One way of addressing this problem is for experienced researchers to organize the information and resources they use into a web portal. The BioAfrica website1 was developed as a web portal and online resource centre for HIV genetic and protein bioinformatics analysis, which gives a special focus on the explosive HIV-1 subtype C epidemic in Southern Africa. Development of the BioAfrica website was started in 2001 by Tulio de Oliveira and Sharon Cassol on the Africa Centre for Health and Population Studies in South Africa in order to manipulate and distribute data produced by their research group. Since then, the website has grown both in size and functionality and presently contains many useful resources and much relevant information for HIV sequence analysis. For example, the website contains information on the distribution of HIV-1 subtypes in Africa and distributes the integrated sequence analysis system for HIV-1 based on the Genetic Data Environment (GDE)2 software. It also hosts automated virus subtyping tool, the REGA HIV, HCV, HBV and HTLV-1 subtyping tools3, and the first web-based HIV proteomics and structural genomics resource worldwide4. The website also points to representative sequences databases, such as the Los Alamos HIV Sequence database5, and mirrors the Stanford HIV Drug Resistance Database6 and the RNA Virus Database7. In addition, the BioAfrica website also points to representative sequences, programs and specific bioinformatics resources that are used for research on the genetic evolution and molecular epidemiology of the local African epidemic as well as antiretroviral drug resistance database. In summary, the BioAfrica website is a well established resource developed to target the emerging need for integrated bioinfor-

The Annotation of Proteins from Pathogens in UniProtKB/SwissProt: Current Status and Future Plans Amos Bairoch Swiss Bioinformatics, Geneva, Switzerland Email: [email protected] The Swiss-Prot knowledgebase (Bairoch et al. Briefings Bioinform 2004;5:39–55. ) was created in 1986. It is now the cornerstone of the UniProt consortium (Wu et al. NAR 34:D187–191, 2006) efforts. We will briefly describe what Swiss-Prot has to offer to the life scientists. The major part of the presentation will be devoted to our current efforts in annotating proteins originating from microbial and viral pathogenic organisms. We will also describe our plan to initiate a concerted effort for a distributed annotation project concerning proteins originating from eukaryotic pathogens. Computing Integrated Genetic Epidemiology Michel Tibayrenc IRD Representative Office, French Embassy 29, Thanon Sathorn Tai, Bangkok 10120, Thailand Email: [email protected] Genetic epidemiology considers the role played by genetic factors in the transmission and severity of infectious diseases. However, specialists of this field usually consider only the host (generally humans). Now, pathogens’ genetic diversity is generally considerable and has a strong impact on the epidemiology of transmissible diseases, including resistance to antibiotics and other drugs. Moreover, the genetic diversity of pathogens can be used for epidemiological tracking (strain typing). Last, in the case of vector-borne diseases, vectors’ genetics also needs to be explored to characterise their populations of vectors, as well as their vectorial capacity and their resistance to insecticides. The integrated approach proposed by us consists in exploring together the impact, on the transmission and severity of infectious diseases, of the host’s, the pathogen’s and the vector’s genetic diversity. As a matter of fact, these three parameters actually correspond to a unique biological phenomenon (coevolution). Genetics is here taken in a broad sense and includes genomics, proteomics, population biology and evolution.

370

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

This approach is the scope of the MeeGID (Molecular epidemiology and evolutionary Genetics of Infectious Diseases) congresses and of the journal Infection, Genetics and Evolution. By taking the example of American trypanosomiasis (Chagas disease), I will illustrate the difficulties of this approach and will discuss the possibilities to apply it to other infectious diseases, including African trypanosomiasis. Information Superstructure for Protozoan Aquaporins Raphael D. Isokpehi1*, N. Chambwe1, J. M. Murray1, H.H.P. Cohly1, S. Varadharajan2, R.V. Rajnarayanan3 1 Department of Biology and RCMI Centre for Environmental Health, Jackson State University, P.O. Box 18540, Jackson, MS 39217, USA; 2 Department of Chemistry, D.G. Vaishnav College, Tamil Nadu, India; 3 Department of Chemistry, Tougaloo College, Jackson, MS, USA * Corresponding author. Email: [email protected] Protozoan parasites of the genera Plasmodium, Trypanosoma and Leishmania contribute significantly to the burden of global infectious diseases. Genome sequencing projects on protozoan parasites combined with high-throughput gene and protein expression experiments are producing wealth of datasets for the discovery of novel drug targets. Water channel proteins, collectively termed aquaporins, are increasingly recognised as potential drug targets or transport routes for anti-protozoan drugs. We are developing an information superstructure on protozoan aquaporins by integrating heterogeneous datasets including genomic sequences, gene and protein expression profiles, protein structures and literature. Conventional aquaporins are six-transmembrane proteins with two AsparagineProline-Alanine (NPA) motifs found in the inter-transmembrane loops B and E that form an aqueous pore. There are 13 known human aquaporins (AQP0–AQP12) while in sequenced protozoan genomes the abundance of aquaporin sequences range from 0 (Cryptosporidium and Theileria) to 7 (Trypanosoma cruzi). Protozoan aquaporins have been shown to also transport nutrients and metabolites between host and parasite that are crucial for initiating an infection and survival during parasite’s life cycle. We report the generation of motif signatures from multiple sequence alignment of 26 aquaporins from seven kinetoplastids (L. major, L. infantum, L. braziliensis, T. brucei, T. congolense, T. cruzi, and T. vivax) and the human aquaporins. These signatures are being examined as tools for aquaporin data integration and prioritization of binding sites for drug design/development. Acknowledgements: National Institutes of Health (NCRR G12RR13459-09) and National Science Foundation (EPS-0556308) In Silico Drug Design Mohammad Afshar Ariana Pharma–Pasteur Biotop, Paris, France Email: [email protected] The ever increasing need for high quality lead molecules against a growing number of targets is a challenge. The emphasis on quality requires hits that are chemically amenable to modifications and that have a good physico-chemical as well as specificity profile. For many targets, the combination of molecular docking against NMR and X-ray 3D structures with in silico ADME & Tox profiling can advantageously replace High Throughput Screening for quality hit identification and further speed the optimization process. This approach enables multiobjective lead optimization, where both potency and the pharmacokinetic profile are improved. Key developments in this area include

the availability of large lead-like and diverse screening libraries, validated docking methods, improving in silico ADME prediction methods, and integrated experimental screening and crystallography. The current methods build on a vast and growing array of experimental data (physico-chemical properties, activity, 3D structures, etc.). This combination presents an unprecedented opportunity for a large number of laboratories to start drug discovery programs without the need to initially invest in expensive laboratory equipment. The author will present his experience of large scale docking campaigns where several million molecules have been screened and hits progressed towards clinical candidates. In Silico Strategies for Target Discovery Eric Mare´chal Comparative Chemogenomics Team, UMR 5168, Institut de recherche en Sciences et Technologies pour le Vivant, CEA Grenoble, France Email: [email protected] In medical sciences, and in widest terms in biology, a target is a broad concept to qualify a biological entity and/or a biological phenomenon, on which one aims to act as part of a therapeutical intervention. Intervention on the target depends on the level of knowledge of the disease and the cultural background of the physician (traditional medicine will focus more on the symptoms in order to end the symptoms; allopathic (western, orthodox) medicine will focus on the causes (i.e. the etiology) in order to oppose the causes of the diseases; homeopathic medicine (in the sense of non-allopathic medicine) will focus on the etiology in order to treat likes with likes; in the widest sense of the word, vaccination is a category of homeopathic treatment). Intervention also relies on the availability of techniques and technologies, e.g. without exogenous agents (surgery, diets), with exogenous biochemicals (natural substances based on traditional pharmacopoeia, drugs derived from traditional pharmacopoeia or from synthetic chemistry, vaccines) or with exogenous genes (gene therapy). Based on an example of a target/drug discovery project undergone in our laboratory to introduce new antimalarials, we show how current strategies for the discovery of target for new drugs actively combine in silico and ‘‘wet’’ experiments and analyses. In silico organisation of genomic and post-genomic information complying with knowledge of the disease in etiologic terms, appears as an efficient source of information to help a target discovery project, as long as accurate analysing and mining tools are available. Objects, attributes and data are consistent with biological entities as defined or understood by bench biologists in reductionist terms, i.e. genes, messengers, proteins, solutes, cellular structures, tissues, organs, etc., and their physiological measures and scales, i.e. mass, size, lengths, concentrations. Processes and functions represent biological operations and alterations over time, growth, transformations, movements, and reactions. Accurate integration of data and assessment of the relations between objects and processes are critical. However, the most striking limitations lie in the analytic and mining tools to explore the stored biological information. In the case of malaria, in spite of an unprecedented effort to organise genomic and postgenomic information regarding its causative agents (Plasmodium sp.) and make it accessible to the largest scientific audience, atypical features of the malarial genomes limit the use of conventional analytical tools.Subsequent objective of a target discovery project, which is from target to lead, drug, vaccine or gene therapy, should not be ignored. The majority of the targets reported in the literature did not open to any realistic therapeutical

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

treatment. An important drawback is the cost of drug development. In this concern, future prospects include the integration of biological and chemical information. Birkholtz LM, Bastien O, Wells G., Grando D., Joubert F., Kasam V., Zimmermann M., Ortet P., Jacq N., Saidani N., Roy S., Hofmann-Apitius M., Breton V., Louw A.I., Mare´chal E. Integration and mining of malaria molecular, functional and pharmacological data: how far are we from a chemogenomic knowledge space? Malaria J 2006;5(1):110.

Speakers Comparative Analysis of Microbial Genomes to Study Expanded Gene Families in Mycobacterium tuberculosis Nicola Mulder*, H. Rabiu, G. Jamieson, V. Vuppu NBN Node IIDMM, University of Cape Town Medical School, Anzio Road, Observatory 7925, South Africa* Corresponding author. Email: [email protected] Mycobacterium tuberculosis, the causative agent of tuberculosis, is the leading infectious disease agent, causing millions of deaths annually. The incidence of disease is increasing with the AIDS pandemic, and current vaccines and therapies are not 100% efficient, resulting in the emergence of drug resistance. The ability of the organism to evolve resistance to drugs with enhanced pathogenicity appears, at least in part, to be provided by the mechanism of gene duplication. This evolutionary mechanism results in expansion of gene families, thereby providing the organism with extra copies of the gene and thus the opportunity to evolve new functions. This project aimed to identify the expanded gene families in M. tuberculosis and investigate the potential contribution of gene duplication events to pathogenicity. Comparative genomics tools were used to compare the M. tuberculosis proteome to itself and the full protein complement from over 80 other microorganisms, including both pathogens and nonpathogens, to determine the extent of family expansion in each. We investigated the link between gene duplication, genome size and GC content, and selected expanded families from M. tuberculosis that were unique to either M. tuberculosis or pathogens for further analysis. Up to half of all M. tuberculosis proteins belong to expanded families or gene duplicate sets, some of which are unique to this organism or to M. tuberculosis complex organisms, suggesting that they have a role to play in evolution of these genomes. Although the evolution of M. tuberculosis is thought to be relatively recent, the maintenance of these duplicated families in the genome suggests they have a role to play in the pathogenic lifestyle of the organism. VectorBase, Resource Center for Invertebrate Vectors of Human Pathogens Karyn Me´gy1*, P. Aarensburger2, P. Atkinson2, N. Besansky3, R. Bruggner3, R. Butler3, K. Campbell4, G. Christophides5, S. Christley3, E. Dialynas6, D. Emmert4, M. Hammond1, C. Hill7, R. Kennedy3, D. Lawson1, N. Lobo3, G. Madey3, R. Maccallum5, S. Redmond5, S. Russo4, D. Severson3, E.O. Stinson3, P. Topalis6, E. Birney1, W. Gelbart4, C. Louis6,8, F. Kafatos5 and F. Collins3 1 EMBL Outstation, European Bioinformatics Institute (EMBL-EBI), Wellcome Trust Genome Campus, Hinxton, CB10 1SD, UK; 2Departement of Entomology, University of California, Riverside, 900 University Avenue, Riverside, CA 92521, USA; 3Center for Global Health and Infectious Diseases, University of Notre Dame, Notre Dame IN 46656– 0369, USA; 4The Biological Laboratories, 16 Divinity Avenue, Harvard University, Cambridge, MA 02138, USA; 5Cell and Molecular Biology Department, Imperial College London, South Kensington Campus,

371

London SW7 2AZ, UK; 6Institute of Molecular Biology and Biotechnology, FORTH, Vassilika Vouton, P.O. Box 1385, Heraklion, Crete, Greece; 7 Department of Entomology, Purdue University, West Lafayette, IN 47907, USA; 8University of Crete, Heraklion, Crete, Greece * Corresponding author. Email: [email protected] Until recently, the malaria vector Anopheles gambiae was the only sequenced and annotated mosquito. But in 2006, the genome of Aedes aegypti, the yellow fever vector mosquito, has been released and annotated. Complementing the gene sets, a large amount of data was available: expression and comparison data, domain content information (signal peptide, transmembrane and protein domains) and control vocabulary. With the upcoming sequencing of more disease vectors (Culex pipiens, Ixodes scapularis, Glossina morsitans, etc.), it is becoming important to organise the storing and access to this data. VectorBase is a resource centre for invertebrate vectors of human pathogens and regroups information about these organisms: gene sets and related information, sequences (genomic, ESTs, traces) and pictures. But more than storing, VectorBase is a place for mining data: walk along the genomes, visualise DNA and protein similarities between organisms, compare sequences vs. the hosted genomic sequences (Blast) and build alignments (ClustalW). VectorBase aims to be a main resource centre for the invertebrates disease vector communities, involving the scientists, generating, updating and giving an easy access to the data [http://www.vectorbase.org/index.php]. Characterisation, Genome Organisation and Phylogenetic Analysis of Two New Viruses Infecting Legume Crops in Ethiopia Adane Abraham1*, M. Varrelmann2 and H. J. Vetten3 1 Ethiopian Institute of Agricultural Research, National Plant Protection Research Center; 2University of Goettingen, Institute for Plant Pathology and Protection, Goettingen, Germany; 3Federal Biological Research Centre for Agriculture and Forestry, Institute for Plant Virology, Microbiology and Biosafety, Messeweg, Germany * Corresponding author. Email: [email protected] While attempting to investigate the viruses associated with yellowing and stunting diseases of cool season food legumes in Ethiopia, two new viruses were discovered and characterised. A distinct luteovirus tentatively named as Chickpea chlorotic stunt virus (CpCSV) was found to be associated with most legume crop samples namely faba bean, chickpea, lentil, grass pea and fenugreek collected from Ethiopia. Biological studies indicated that CpCSV is transmitted by the aphid Aphis craccivora but not by A. fabae, Acyrthosiphon pisum or Myzus persicae and has an experimental host range limited to species of cool season food legumes. Electron microscopy showed isometric particles of 28 nm in diameter typical of luteoviruses. A rabbit polyclonal antiserum and 10 mouse monoclonal antibodies were produced, characterised and evaluated for routine use in its diagnosis. Sequencing of the complete genome of an Ethiopian (Ambo) isolate revealed that its linear RNA genome has 5900 nucleotides arranged in six major open reading frames typical to members of genus Polerovirus in the family Luteoviridae. Phylogenetic analysis of the polymerase and coat protein (CP) sequence using Clustal-X program confirmed this classification. Recombination analysis of the genome using Sister scanning procedure suggested that although the virus genome is generally close to Cucurbit aphid-borne yellows virus, it has acquired a stretch of about 90 amino acids at the C-terminal part of its readthrough domain from ancestral soybean dwarf virus-like ancestor. The CP protein gene sequence of 18 CpCSV

372

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

isolates from five countries indicated that the virus occurs as two geographically differentiated strains significantly differing in symptom severity, serological properties and nucleotide sequence information. Serological analysis of 73 nanovirus isolates from faba bean samples collected from Ethiopia indicated that the isolates could be categorised into at least three distinct serotypes designated A, B and C. The nucleotide sequence of at least two genes encoding for the CP and U1 protein were obtained for representative samples of each serotype. Whereas serotype A and B isolates were similar to those reported elsewhere previously, Serotype C represented a new group. To understand the phylogeny of the later in detail, the complete sequence of the eight distinct circular single stranded DNAs believed to form entire nanovirus genome was obtained from a representative sample (eth-231). Each of these DNAs was about 1 kb in size and code for proteins of 12–35 kDa. Sequence comparison indicated that overall nucleotide similarity of only 70% with Faba bean necrotic yellows virus, its closest relative, suggesting that it is a distinct nanovirus belonging to family Nanoviridae. For this apparently new virus, a tentative name faba bean yellow leaf virus (FBYLV) is proposed. DNA Barcoding and Taxonomy of Glossina spp. Daniel Masiga1* and J. Ouma2 1 Molecular Biology and Biotechnology Department, ICIPE, P.O. Box 30772–00100 Nairobi, Kenya; 2Trypanosomiasis Research Centre, Kenya Agricultural Research Institute, P.O. Box 362, Kikuyu, Kenya * Corresponding author. Email: [email protected] The study of the broad field of biology relies extensively on the ability to classify and identify individuals to the lowest taxonomic level. It is estimated that 250 years of dedicated taxonomic activity has resulted in an impressive 1.5–1.8 million descriptions, representing only about 2% of all species, estimated at 5–100 million. Unfortunately, expertise in taxonomy based entirely on morphology is becoming more difficult to find, and largely resident in national life science museums. Identifying species accurately is foundational to the ecology of insects that spread disease, and the pathogens they transmit. Although application of unique DNA sequences is not new to species identification, the application of a standardised locus for implementation on a broad scale is, and continues to elicit much debate. The Consortium for the Barcode of Life (CBOL, www.barcoding.si.edu) has identified mitochondrial Cytochrome Oxidase I (COI) as the barcode sequence for animals. Using universal primers described by Folmer et al. (1994), the barcoding region (as defined by CBOL) we have obtained amplicons from six species of tsetse (G. fuscipes fuscipes, G. pallidipes, G. swynnetoni, G. morsitans morsitans, G. morsitans centralis and G. m. submorsitans). Under our PCR conditions, G. brevipalpis and G. longipennis failed to amplify, hence more optimisation is needed. However, all the six produced high quality sequences in excess of 650 bp from the PCR products. These data demonstrate the utility of COI in identifying tsetse flies to the level of sub-species. A Convoluted Problem and Heuristic Solution for Predicting Ensembles and Transitions for Pseudoknots Polymers Ezekiel Adebiyi Department of Computer and Information Sciences, Covenant University, PMB 1023, Ota, Nigeria Email: [email protected]

Note that tRNAs provides potential drug targets and are readily available. Presently, little is known about the structures of tRNA in Plasmodium falciparum, so development of computational techniques toward the prediction of tRNA structures will enhance the work on structural based drug design for malaria treatments. Chen and Dill (PNAS, 97, 646–51, 2000) developed a method capable of predicting structures, but its can not work on tRNAs. Lucas and Dill (J Chem Phys 2003;119(4), 2414–21) and Kopeikin and Chen (J. Chem. Phys. 2005;122(9), 1–13) obtained one independently but their techniques worked only for simple structures. In fact, the prediction of Lucas and Dill overestimate in its prediction. In the work that follows, we took a broader look at a larger class of structures and developed a heuristic method for the resulting convoluted problem. Mainly, our technique targets extending the theory on polymer graphs that exist (PNAS, 97, 646–651, 2000) to include pseudoknots. Our experimentation produced enhanced results. Our future work focused on the improvement of our present framework. ˆ te d’Ivoire: Clade Diversity of HIV-1 Subtype A Pol Gene in Co Evidence of Underestimated Rate of Recombinant Form Gonedele´ Bi Sery1,2*, S. Abdourahmane1, C. Gueladio2, F. Gnangbe´.1, N. Eliezer2,3 1 De´partement de ge´ne´tique, Universite´ d’Abidjan-Cocody, 22 BP 582 Abidjan 22, Coˆte d’Ivoire; 2Centre Suisse de recherches Scientifiques en Coˆte d’Ivoire, 01 BP 1303 Abidjan 01, Coˆte d’Ivoire; 3De´partement de parasitologie, Universite´ d’Abidjan-Cocody, 22 BP 582 Abidjan 22, Coˆte d’Ivoire * Corresponding author. Email: [email protected] Based on partial env and pol (protease and RT) subtyping, it was recently documented that the majority (>80%) of the HIV-1 strains that circulate in Coˆte d’Ivoire were CRF02_AG and about 11% could not be clearly assigned to a known subtype or Crf (circulating recombinant form). Thus many subtypes A sequences from this country submitted to HIV genome databases are still not assigned to any sub-subtype. So the number of subtype A sequences not assigned to any sub-subtype continues to grow. Inter- and intra-clade variations within pol sequence are particularly relevant as this region encodes RT and protease proteins, against which many antiviral drugs are directed. We attempted to identify and characterise HIV-1 subtype A using all pol gene sequences of this subtype isolated from Coˆte d’Ivoire and registered in the NCBI GenBank and Los Alamos database (n = 85). Rega was used for subtype identification and Simplot was used for Bootscan analysis to detect recombinant forms. A maximum-likelihood tree of pol gene sequences was produced by DNAML program in the PHYLIP package version 3.6 using a transition/transversion ratio of 1.5 and empirical base frequencies. Sequences were gap stripped. Pairwise genetic distances were calculated by using the DNADIST program in PHYLIP with an F84 model of evolution and a transition/ transversion ratio previously defined. Calculations were done with the reference sequences A1, A2, A3, and AG. All the clade previously defined as subtype A clustered with the recombinant forms CRF02_AG and AG. Distance calculations between these sequences and reference sub-subtypes A1, A2, A3 (7– 11.5%) and CRF02_AG (1.7–3.8%) fell in the range of distances previously characterised between and within sub-subtype groups. Our results provide insights into the importance of recombinant forms in Coˆte d’Ivoire which seems to have been underestimated,

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

and thus indicate that more than 90% of HIV-1 strains are recombinant forms. Analysis of Signatures of Polymorphisms and Positive Adaptive Selection at Candidate Genes for Genetic Resistance to Avian Viral Diseases Sheila Cecily Ommeh1,2*, D. Masiga1, D. Lynn3, H. Jianlin2, O. Hanotte2 1 Department of Biochemistry and Biotechnology, Kenyatta University, P.O. Box 43844, Nairobi, Kenya; 2International Livestock research Institute (ILRI), P.O. Box 30709, Nairobi, Kenya; 3Smurfit Institute of Genetics, Trinity College Dublin, Dublin 2, Ireland * Corresponding author. Email: [email protected] (P.O. BOX 52856 CSQ 00200 Nairobi, Kenya) The objective of this study is to assess polymorphisms at candidate genes associated with genetic resistance to viral diseases in chicken, such as Marek’s disease and avian Influenza. Two candidate genes were shortlisted from literature and their sequences obtained from Ensembl. BLB2 (class II) major gene is present in chicken within the MHC region on chromosome 16. It has been shown to be associated with resistance to avian viral infection, including Marek’s disease. Mx gene on chromosome 1 is a putative candidate gene for genetic resistance to avian Influenza at position 631 of the encoded amino acid. Species homologs of these two genes were obtained from the NCBI Blast and analysed with the PAML package for detection of signature of positive adaptive selection (dN/dS ratio). Lineage analysis of species homologs of the Mx gene revealed a dN/dS ratio of 0.64 implying purifying selection, but site analysis revealed five amino acids under positive selection. These sites under positive selection lie in the central interaction domain CID of the Mx gene that mediates the activities of the GTPase domain and the leucine zipper domain which is important in the oligomerisation of the Mx protein. For the BLB2 gene lineage analysis of species homologs did not reveal positive selection but site analysis revealed seven amino acids under positive selection which lie in the peptide binding region PBr of the protein. This region is encoded by exon 2 of the BLB2 which is important in antigen presentation. In summary site analysis reveals that some amino acids in the CID and PBr domains for the Mx and BLB2 genes, respectively, are under positive selection and this warrants further research on these two protein domains for their exact role towards genetic disease resistance. A Molecular Integration Database System for All: Integration of HIV Clinical and Molecular Data Using Open Source Genome Management Tools Heikki Lehvaslaiho1*, Adam Dawe1*, Allan Kamau 1, Ruby van Rooyen1, Tulio de Oliveira1, Anelda Boardman1, Alan Powell1, Salim Abdool Karim2, Koleka Mlisana2, Lynn Morris3, Clive Gray3, Carolyn Williamson4 and Winston Hide1 1 SANBI, UWC, Bellville, South Africa; 2Centre for AIDS Programme of Research in South Africa (CAPRISA), Nelson R Mandela, School of Medicine, Durban, South Africa; 3National Institute for Communicable Diseases, Johannesburg, South Africa; 4Institute for Infectious Diseases and Molecular Medicine, University of Cape Town, South Africa * Corresponding author. Email: [email protected] HIV studies can combine long term clinical monitoring, sequencing of large number of whole or partial viral genomes to assess viral diversity, multi-institutional molecular laboratory assays including immune monitoring, and host genetics. Integrating this form of data

373

across molecular and non-molecular domains is a challenge. In order to maximise the rate at which discoveries can be achieved in our study, we have developed a strategy of leveraging global open source developments to apply and refine in a developing world context. Capturing contextual information together with time series genome sequences from infected subjects is non-trivial. As part of an nIH acute infection study of the Centre for AIDS Programme for research in South Africa, we have developed a Molecular Integration Database (MID). It securely and reliably integrates clinical, immunological and molecular data from infected subjects. The MID leverages open source software developed for model organism genomes such as BioMart for querying and the Generic Genomic Browser (GBrowse) for visualisation. Use of genome project software means that the resultant saving of time and resources in the database’s development has been substantial and in return, the developing world, where infection is endemic, and its open source community will benefit from the release of the source code of the MID and its HIV sequence assembly pipeline. GONNA: A Gene Ontology Nearest Neighbour Approach for the Functional Prediction of Plasmodium falciparum Orphan Genes: The Database of the Predictions Laurent Bre´he´lin*, J.F. Dufayard, O. Gascuel Me´thodes et algorithmes pour la Bioinformatique, LIRMM, CNRS Univ. Montpellier II, 161 rue Ada, 34392 Montpellier Cedex 5, France * Corresponding author. Email: [email protected] Plasmodium falciparum, the pathogenic agent responsible for malaria, causes close to 3 millions human deaths each year in the world. Its genome, published in 2002, remains poorly understood. Among its 5000 genes, 3000 have a totally unknown biological function. Several reasons explain this situation, the first one being a totally atypical genome composition which renders ineffective the usual methods of functional annotation based on sequence comparison (alignment) with homologous genes in nearby organisms. Nonhomology methods are thus needed to obtain functional clues for the orphan genes. Notably, transcriptomic analysis using DNA microarrays have been proposed. These approaches are based on the hypothesis that uncharacterised genes may potentially share functional roles with annotated genes of similar transcriptomic profile, a principle known as ‘‘Guilt by Association’’ (GBA). Interestingly, this intra-species principle can also be apply to other type of available postgenomic data, such as proteomic data or protein interaction data. The GO Consortium (http://www.geneontology.org) has developed a systematic and standardised nomenclature for annotating genes in various organisms, including P. falciparum. The GO annotation is a hierarchical structure describing generalisation relationships between hundreds of terms. So each gene can have several associated GO terms, due to the hierarchical structure of the ontology, and of the existence of multi-functional genes. Here, we propose a low computing time method that allows the application of the GBA principle on the Gene Ontology (GO) in a very extensive way, and for several postgenomic data. Notably, it allows intensive used of a ‘‘cross-validation’’ procedure to provide high quality assessment of the predictions of the method for each GO term individually. The result is an assessed functional draft of the P. falciparum orphan genes. This draft has been integrated in a database available on the web and that can be accessed in different ways. A global view allows to rapidly inspect the branches of the ontology that are more suitable for the GBA principle using the considered type of data (transcriptome, proteome, and protein interaction data). Moreover, user can query the database to search

374

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

for the potential GO terms of one particular gene, or to search for all the genes that potentially belong to a given GO term. Web site: http://atgc.lirmm.fr/plasmo_draft/ Global Analysis of Leishmania Genes Expression Using SAGE Libraries Sondos Smandi1, F. Guerfali1, D. Mallouche2, K. Ben-Aissa1, O. Mghirbi1, L. Guizani1, D. Laouini1, A. Benkahla1* and K. Dellagi1 1 LIVGM, Institut Pasteur, Tunis, Tunisia; 2LEGI-EPT, Ecole Supe´rieure de la Statistique et de l’Analyse de l’Information de Tunis, Tunisia * Corresponding author. Email: [email protected] Leishmania is a major parasitic disease in the world for which no vaccine is available and drugs are toxic. As part of its research programme on host-parasite interaction, the LIVGM (Laboratoire d’Immunologie Vaccinologie et Ge´ne´tique Mole´culaire) has constructed four human macrophage SAGE libraries, either non-infected or infected by Leishmania, and one Leishmania promastigote library. The aim of this ‘‘SAGE project’’ is to dissect the impact of intracellular infection on the expression of macrophage genes. The aim of the present project is to identify and classify Leishmania derived tags, to evaluate the expression of Leishmania genes and to identify those upor down-regulated in the amastigote stage. The challenge is due to the fact that: (i) very few full length Leishmania cDNAs are available in public databases; (ii) the only Leishmania gene evidences are the CDS predictions available in GeneDB (www.genedb.org); and (iii) tags are normally located in the 30 -UTR (outside the CDS). We will present here the strategy followed in order to assign Leishmania tags to the corresponding transcripts/gene and to analyse their expression level: (i) Leishmania tags mapping was done using Blast; (ii) the association tag-to-gene was done using Gaussian kernels; (iii) the analysis of differential expression was done using the binomial distribution. Multi-SOM: A Novel Clustering Approach for Gene Expression Analysis Amel Ghouila1,3*, H. Jmel3, S.B. Yahia1, D. Malouche2, S. Abdelhak3 1 Computer Science Department, faculty of Sciences of Tunis, Tunisia; 2 LEGI-EPT-ESSAI, 7 November University, Carthage, Tunisia; 3Pasteur Institute of Tunis, Tunisia * Corresponding author. Email: [email protected] The production of increasingly reliable and accessible expression data has stimulated the development of computational tools to interpret such data and to organise them efficiently. The clustering techniques are largely recognised as useful exploratory tools for gene expression data analysis. Genes that show similar expression patterns over a wide range of experimental conditions can be clustered together. This relies on the hypothesis that genes that belong to the same cluster are co-regulated and involved in related functions. Clustering involves working with a data matrix in which the rows represent the genes, the columns represent the experiments, and each entry represents the expression level of a gene in a given experiment. There exist many clustering algorithms which take microarray data sets as input and produce clusters as output. In spite of the fact that a variety of different clustering algorithms is now available, a number of important questions remain to be addressed. Especially, the estimation of the number of clusters represented in an expression data set is a crucial and complex task, which may significantly influence the outputs of an analysis process.

We propose here a multi-level SOM based clustering algorithm titled Multi-SOM. Through the use of clustering validity indices, Multi-SOM overcomes the problem of the estimation of clusters number. It tends to find the optimal clustering of data. To test the validity of the proposed clustering algorithm, we first tested it on supervised training data sets. We evaluate our results by computing the number of misclassified samples. We have then validated Multi-SOM on publicly available biological data. Finally, we discuss the use of MultiSOM on microarray data of gene expression in the native skin compared to skin derivative cells and on macrophage data infected with different pathogens. Gene ontology tools are then used to uncover the properties shared among, and specific to, clusters of genes produced by applying Multi-SOM algorithm. To better visualise and analyse these clusters, we used the web based software, GOTM, FatiGo, ONTO-Express and GoSurfer. These tools are very useful to look for the common traits that are shared within gene clusters, since terms are organised in three general categories, ‘‘biological process’’, ‘‘molecular function’’, ‘‘cellular process’’. To show the specificity of each cluster, statistical tests are also performed. Dynamics of Migrations of Malaria Protective Polymorphisms across Africa’s Sahel Niven Abdel Wahab Salih*, H.Y. Hassan, A. Hussein, D. Kwiatkowski and M.E. Ibrahim Institute of Endemic Diseases, University of Khartoum, Sudan * Corresponding author. Email: [email protected] The Sahel which extends from the Atlantic Ocean to the Ethiopian highland is a historical reservoir of Africa cultures and grandest populations and a known arena of ancient and recent migrations. We study polymorphisms of mitochondrial and Y chromosome haplotype and SNPs of known protective effects against malaria in populations that lives across the Sahel, e.g. Hausa, Massalit, Bergu as well as other groups extending into the eastern Sudan. Populations were classified according to their linguistic affiliation which comprised the main linguistic families of Africa (Afro-Asiatic, Niger Kordofanian and Nilo-Saharan). We present evidence that despite the known historical endemicty of malaria in the eastern Sahel, some of the major protective polymorphisms in malaria have in fact found their way only recently to the gene pool of the populations in eastern Sahel. We give estimates of the age of the introduction of such alleles and discuss the possible dynamics of the process and its impact on the epidemiology of the diseases. Identification of Theileria parva Vaccine Candidate Genes Using a Bioinformatics Approach Etienne de Villiers International Livestock Research Institute, P.O. Box 30709–00100, Nairobi, Kenya Email: [email protected] Immunity against the bovine protozoan parasite Theileria parva has previously been shown to be mediated through lysis of parasiteinfected cells by MHC class I restricted CD8+ cytotoxic T lymphocytes. It is hypothesised that identification of CTL target schizont antigens will aid the development of a sub-unit vaccine. We exploited the availability of the complete genome sequence data and bioinformatics tools to identify genes encoding secreted or membrane anchored proteins that may be processed and presented by the

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

MHC class I molecules of infected cells to CTL. A 525 base pair Orf encoding a 174 amino acid protein, designated Tp2, was identified by T. parva- specific CTL from four animals. These CTL recognised and lysed Tp2 transfected skin fibroblasts and recognised four distinct epitopes. Significantly, Tp2 specific CD8+ T cell responses were observed during the protective immune response against sporozoite challenge. This makes Tp2 an attractive candidate for evaluation of its vaccine potential. Discovery of Novel Drug Targets Against Pathogenic Protozoa: The Promise of Metabolic Reconstruction Lillian Wambua1,2, G. Mcconkey1, D.R. Westhead1* 1 Leeds University, School of Biological Sciences, Bioinformatics Department, LS2 9JT, Leeds, UK; 2International Livestock research Institute, Nairobi, Kenya * Corresponding author. Email: [email protected] Protozoan organisms as Plasmodium, Toxoplasma, Eimeria, Babesia, Theileria, Trypanosoma, Leishmania, Cryptosporidium and Entamoeba species require intensive research for discovery of new drugs to circumvent drug resistance. The availability of complete genome sequences for many protozoans has set stage for mining of their genomes through systems biology approaches for novel drug targets. The metabolic architecture of the organisms in particular provides a rich resource for potential targets. We sought to identify presence of unique metabolic enzymes and pathways that would make possible drug targets in the protozoans of interest by exploring their metabolism through online database searches. However, profound under-representation of these organisms on online public databases rendered this approach unfruitful. We therefore performed metabolic reconstruction of their metabolic pathways from their genome sequences using metaSHArk, a metabolic reconstruction tool developed by the Leeds University. This software combines a gene-detection package which identifies metabolic enzymes within raw un-annotated DNA sequences or ESTs. Gene detection is based on a two-stage protocol involving a PSI-TBLASTN search on the PRIAM database followed by application of Wise2 algorithm with Hidden Markov Models to assert the enzymes. We studied five metabolic pathways: Glycolysis, Krebs’s cycle, Shikimate, Pantothenate, Coenzyme A and Folate biosynthetic pathways. Our results showed potential targets in shikimate and folate pathways in Plasmodium, Toxoplasma and Eimeria species. Babesia, Theileria and Plasmodium had highly similar metabolic profiles, suggesting prospects of using anti-malarial drugs for treatment of Babesiosis and Theileriosis. Toxoplasma and Eimeria species had robust metabolic capacities compared to all other protozoans explored and therefore had several potential drug targets in the folate, pantothenate and shikimate metabolic pathways. Cryptosporidium and Entamoeba, have drastically reduced metabolic repertoire. However, hope lies in the mining of their genomes for unique bacterial genes acquired by possible lateral gene transfer. Trypanosoma and Leishmania species also showed reduced metabolic capacities, apart from their elaborate carbohydrate metabolism pathways. The evolutionary compartmentalisation of their glycolysis pathway in specialised organelles referred to as glycosomes may have rendered these enzymes distinct from those of the human host, thereby making them potential drug targets. This is subject to further comparative genomic analysis. We found the metabolic capacities of the organisms to be intricately associated with the endosymbiotic evolutionary origins of the organelles where these pathways were localised in the cell,

375

most notably the mitochondria and the apicoplast. Loss of these organelles through secondary events was congruent with the absence of these pathways in the organism. LBTx: A Tool Box for the Selection and Analysis of Potential Leishmania Target for Intervention and Disease Control Ikram Guizani*, A.S. Chakroun, L. Turki, H. Kebaier, M. Barhoumi, Y. Saadi, I. Smeti, A. Sahli, S. Guerbouj, R. Lahmadi, M. Ben Fadhel, M. Driss, B. Kaabi, K. Ajroud, J. Chemkhi, S. Ben Abderrazak Laboratoire d’Epide´miologie et d’Ecologie Parasitaire, Institut Pasteur de Tunis, Tunisia * Corresponding author. Email: [email protected] Protozoan parasites of the genus Leishmania are responsible for leishmaniases, neglected diseases having a wide geographical distribution. These highly prevalent diseases cover a wide range of clinical presentations and constitute major public health problems. In spite of the importance and the number of studies achieved on these diseases, they remain poorly controlled. The identification of new targets/novel approaches for diagnostics, vaccine or treatment remains a research priority. As a part of our work, we developed a bioinformatics based support that we customised to the specific needs of our research approaches. This tool has been implemented to analyse polymorphic DNA markers from different Leishmania parasite species like L. major, L. aethiopica, L. tropica, L. infantum, L. donovani and L. archibaldi, identify polymorphic microsatellite loci, and design of specific PCR primers using a comparative genomic approach. It has also been instrumental for the comparative genome analysis and annotation of differentially expressed Leishmania transcripts, or Leishmania antigen genes and families. This tool called LBTx for ‘‘Leishmania Bioinformatic’s ToolboX’’ contains interfaced programmes and scripts among which Blast where its searching database became only constituted by public releases of the different Leishmania genomes data. Other scripts were developed in Perl language (MutationCheckUp, pepHasher, MisaCluster). As a matter of example, MisaCluster is a microsatellite viewer and a primers/probes mapper that takes multiple sequence alignments as input data. This tool is used in association with other bioinformatic programs like Artemis, ACT, Clustal, MEME, primer 3, etc. This Toolbox keeps under development and can be expanded to the needs on request. This work received financial support from MeSrST–Tunisia (Contract-programme 2004–2008), from the UnICef/UnDP/World Bank/WHO special program for research and training TDr (A30380, A11032, A30134) and from rAB & GH programme of WHO/eMrOCOMSTeCH. Computational Discovery of Drugs Resistance Mechanism(s) of the Malaria Parasite Plasmodium falciparum Marion Owolabi1, S. Fatumo1,2, R. Koenig2, G. Schramm3, A.-L. Kranz2, R. Eils2,3, E. Adebiyi1* 1 Department of Computer and Information Sciences, Covenant University, PMB 1023, Ota, Ogun State, Nigeria; 2Department of Bioinformatics and Functional Genomics, Institute for Pharmacy and Molecular Biotechnology, University of Heidelberg, 69120 Heidelberg, Germany; 3Theoretical Bioinformatics, German Cancer research Centre (DkfZ), 69120 Heidelberg, Germany * Corresponding author. Email: [email protected] At the genomics level, we use the microarray technology to produce gene expression data for various organisms under many conditions. A

376

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

desirable advancement is the need to extract from this gene expression data, information that will be useful toward given answers to the questions targeted at the design of the microarray experiment. Analysis of gene expression data of Plasmodium falciparum when induced with two anti-malaria drugs (such as chloroquine and choline analogue T4) has shown that the parasite resistance mechanisms may not be elucidate-able at the genomics level. But on a proteomic level, biochemical research has elucidated an increasingly complete image of the metabolic architecture of organisms that included that of P. falciparum. In this work, we sought to use the biochemical network of P. falciparum to deduce its drugs resistance mechanism(s) using the two gene expression data obtained when the parasite is treated with chloroquine and choline analogue T4. We do this by mapping the gene expression data onto the enzymatic reaction nodes of the metabolic network. A consecutive clustering method is used to derived important clusters. Further, a wavelet and a feature extraction method is applied to study these clusters which gives important insight into the mechanisms that P. falciparum deplore for resisting anti-malaria drugs. Large-scale Distributed In Silico Drug Discovery Using VSM-G (Virtual Screening Manager for Computational Grids) Le´o Ghemtio*, B. Maigret, A. Beautrait, V. Leroux, M. Chavent Orpailleur and Algorille teams, LORIA, UMR CNRS 7503, Nancy University, France * Corresponding author. Email: [email protected] The Virtual Screening Manager platform, dedicated to automated virtual high-throughput screening using cluster grids, is presented here. The VSM-G package is constituted of pre-processing engines for both protein targets and small molecules putative ligands to be screened on, and of a funnel docking strategy. The latter is architectured as a sequence of different docking modules ranging from a fast but less accurate surface-matching procedure for crude rigid dockings to slower but more elaborated molecular dynamics calculations. At each step of the funnel and depending on tuning, a small proportion of molecules can be prioritised as more promising, discarding a large body of inappropriate ligands. VSM-G is able to handle several millions of compounds vs. hundreds of targets (which can be different targets or several conformations of same one, e.g. extracted from molecular dynamics sampling or nMr data), yielding to a small set of putative hit compounds to be proposed for real biological testing. The high throughput virtual screening experiment, which will be described, aimed to identify, from a database consisting of about 600,000 drugable molecules, putative hits for the LXr receptors. In order to handle both compounds and receptor flexibilities, 400 conformers were generated for each molecule, leading to a total of about 91 millions of 3D objects to be processed. On the receptor side, four different crystal conformations were used for the docking calculations. Because considerable computing power is needed for performing searches of this type within a reasonable time, VSM-G targets the use of computational grid technology, especially for the fast-matching procedure used in the first selection docking filter. This filter, which had to approximate all molecules and target surfaces using spherical harmonics polynomials, was used to eliminate most of the tested compounds in order to pass only about 2000 molecules to the next flexible docking filter. In order to obtain the molecular surfaces of the 91 M 3D objects, the French GRID’5000 facilities were used adding a total of 1360 processors.

At the end of this experience, we were able to propose a list of only 20,000 molecules to really be tested by our pharmaceutical company partner, reducing therefore the cost for an in vitro HTS campaign while increasing the chances of successes. Several perspectives for expanding VSM-G functionality, some of which are currently developed, will also be discussed. In Silico Docking Against Malaria: The WISDOM Initiative Vinod Kasam, V. Breton*, A. Da Costa LPC Clermont-ferrand, Campus des Ce´zeaux 63177 Aubie`re Cedex, France on behalf of the WISDOM collaboration * Corresponding author. Email: [email protected] The WISDOM initiative aims at developing new drugs for neglected and emerging diseases with a particular focus on malaria. Its specificity is to extensively rely on emerging information technologies to provide new tools and environments for drug discovery and development. Its main goal is to boost research and development on neglected diseases by fostering the use of open source information technology for drug discovery. The WISDOM collaboration is presently using the grid paradigm to identify potential hits in virtual screening. Started in 2005, the initiative has been initially focusing its effort on the haemoglobin metabolism of the malaria vector which is one of its key metabolic processes. Plasmepsin, the aspartic protease of Plasmodium, is responsible for the initial cleavage of human haemoglobin. The goal was to identify which molecules could dock on the protein active sites in order to inhibit its action and therefore interfere with the molecular processes essential for the pathogen. During the summer of 2005, a first large scale deployment allowed achieving in silico docking of 500,000 compounds in about 6 weeks against different structures of plasmepsines II and IV [1]. Results were analysed and the top 1000 compounds were further processed using a Molecular Dynamics procedure to select finally 25 compounds which are now being tested in vitro. With the success achieved by this first initiative both on the computation and biological sides, several scientific groups around the world including groups in Africa proposed targets implicated in malaria which led to a second assault on malaria during the fall of 2006 [2]. WISDOM-II project dealt with several targets which are both X-ray crystal models and homology models. Targets from different classes of proteins were tested: reductases like DHfr as well as transferases such as GST. Analysis of this second initiative is under way. The WISDOM production environment, which was initially designed to allow docking at a very large scale, is now evolving towards a fully grid-enabled virtual screening process where the compounds go through several selection steps using docking and molecular dynamics procedures. In the coming months, we aim at expanding our collaboration with the research communities on neglected diseases to inform them of this service which allow cost-effective screening of their most promising targets. We also foresee to apply our virtual screening process to a subset of the targets proposed by the TDR drug target portfolio network: http://www.who.int/tdr/tropics/ discovery_research/drug_target.htm In a longer term, our vision goes beyond virtual screening and we call for a distributed, internet-based collaboration [3] to address one of the worst plagues of our present world, malaria. The spirit is a non-proprietary peer-production of informationembedding goods. And we propose to use the grid technology to enable such a world wide open-source like collaboration.

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

References [1] N. Jacq et al. Grid enabled virtual screening against malaria. J Grid Comput, in press. [2] J. Salzemann et al. Grid enabled High Throughput Virtual Screening Against four Different Targets Implicated in Malaria, Proceedings of Healthgrid Conference 2007, Studies in Health Technology and Informatics; 2007, 126. [3] V. Breton et al. Grid Added Value to Address Malaria, IEEE Transact NanoBiosci. in press.

Posters The Complete Sequence of the RNA Genome of Sweet Potato Chlorotic Fleck Virus Valentine Aritua1*, E. Barg2, E. Adipala3, H. J. Vetten2 1 National Agricultural Biotechnology Centre, Kawanda Agricultural Research Institute, P.O. Box 7065, Kampala, Uganda; 2Biologische Bundesanstalt fur Land und Forstwirtschaft, Institut fur Pflanzenvirologie, Mikrobiologie und biologische Sicherheit, Messeweg 11–12, D38104, Braunschweig, Germany; 3Department of Crop Science, Makerere University, P.O. Box 7062, K ampala, Uganda * Corresponding author. Email: [email protected] The complete nucleotide sequence of the single-stranded RNA genome of a Ugandan isolate of Sweet potato chlorotic fleck virus (SPCFV) was determined. Excluding the 30 -terminal poly(A) tail the genome is 9104 nucleotides in length. It comprises six open reading frames (ORFs).ORF 1 codes for a 238 kDa protein with characteristic replicase motifs. ORF 2, -3 and -4 code for proteins of 27.5, 11.5 and 7.3 kDa, respectively, shows strong homologies with triple gene block proteins. ORF 5 encodes the capsid protein of 33 kDa and ORF 6 encodes a protein of 15 kDa with a nucleic acid binding zinc finger motif. Based on the genomic organisation SPCFV appears to be a member of the genus Carlavirus with the family Flexiviridae. Comparison of the nucleotide and predicted amino acid sequences of the putative polymerase and coat protein with the members of Flexiviridae also showed highest similarities with the corresponding gene products of carlaviruses. However, SPCFV takes a position close, but clearly separate from the carlaviruses, as none of its putative gene products shared amino acid sequence similarities of >40% with the homologues of other carlaviruses. Based on CP sequences, its closest relative among the carlaviruses is Melon yellowing-associated virus, a proposed carlavirus from Brazil with which it shares 46% aa homology. The larger RN A (9104 nt) of SPCFV in comparison to other carlaviruses (7.4–8.5 kb) is largely due to the considerably larger replicase (238 vs. 200–223 kDa) and the considerably long un-translated region between ORF 4 and the putative CP (213 vs. ca. 34 nt for PVM). Although SPCFV had earlier been regarded as a potyvirus, our data provided unambiguous evidence for its assignment as a distinct species to the genus Carlavirus. In Silico Identification of Candidate Pathogenicity Islands in Complete Xanthomonas Genomes Mtakai Vald Ngara*, K. B. Opap, D.-J. Kim International Institute of Tropical Agriculture (IITA), P.O. Box 30709– 00100, Nairobi, Kenya * Corresponding author. Email: [email protected]

377

The need to integrate computational power and techniques in biological research in the African continent is crucial in tackling disease, food insecurity and managing genetic resources. The International Institute of Tropical Agriculture (IITA) is supporting efforts that will enhance the integration of computation in biological research and exploitation of the numerous bioinformatics resources in research. Currently, IITA is conducting basic in silico research on phytopathogens of the African major food crops. The genus Xanthomonas is a diverse and economically important group of phytopathogens, belonging to the subdivision of the Proteobacteria. The species and pathovars, cause distinct disease phenotypes in a wide range of plants. Pathogenicity islands (PAIs), distinct genomic segments encoding virulence factors, represent a subgroup of genomic islands (GIs) that have been acquired by horizontal gene transfer event. Identification of potential PAI is useful in understanding disease pathogenicity, host specificity and genomic evolution in microbial genomes. In this study, we are utilising an in silico approach to detect potential PAI(s) in six complete Xanthomonas genomes by combining sequence homology and abnormalities in genomic composition. We will first identify the GI(s) in the genomes, by calculating the G + C content anomalies and codon usage bias. All the PAI loci will be collected and their homologs identified in the genomes to come up with genomic strips. An algorithm will then be developed to merge overlapping or adjacent strips into large genomic regions. Among the defined genomic regions, PAI-like regions will be identified by the presence of homolog(s) of virulence genes. Candidate PAI (cPAI) will be considered only if the PAI-like region partly or entirely spans the GI. The identification and analysis of cPAI in Xanthomonas will broaden our knowledge on disease phenotype diversity, host specificity and genomic evolution in these phytopathogenic bacteria. Differential Expression of the CrV1 Haemocyte Inactivation-associated Polydnavirus Gene in the African Maize Stemborer Larvae Parasitized by Two Biotypes of the Endoparasitoid Cotesia sesamiae (Cameron) Catherine Wanjiru Gitau1,4*, D. Gundersen-Rindal2, M. Pedroni2, S. Dupas3, J. P. Mbugi1 1 Kenyatta University, Department of Zoology, School of Pure and Applied Sciences; 2USDA-ARS. Insect Biocontrol Laboratory, Beltsville, MD, 20705, USA; 3IRD, c/o CNRS, Laboratoire Populations Ge´ne´tique et Evolution, 1 av. de la Terrasse, 91198 Gif-sur-Yvette, France; 4ICIPE, P.O. Box 30772–00100, Nairobi, Kenya * Corresponding author. Email: [email protected] The braconid Cotesia sesamiae is the most common endoparasitoid of the cereal stem borers Busseola fusca and Sesamia calamistis in subSaharan Africa. In Kenya, two biotypes of C. sesamiae exist. The C. sesamiae population from the western highlands completes development in B. fusca whereas eggs by C. sesamiae from the coastal population get encapsulated and ultimately, no parasitoids emerge from the B. fusca larvae. Both populations successfully complete development in S. calamistis larvae. Results from this study showed that B. fusca larvae parasitised by the avirulent C. sesamiae had frequently more encapsulated eggs at 24 h post parasitism compared to earlier time points. Expression of the C. sesamiae polydnavirus (PDV) CrV1 gene associated with haemocyte inactivation in the Cotesia rubecula/Pieris rapae system was detected in fat body and haemolymph samples from both B. fusca and S. calamistis larvae parasitised by the virulent C. sesamiae from Kitale and Meru regions. There were no or faint amplified CrV1 products on B. fusca fat body

378

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

and haemolymph samples parasitised by the avirulent C. sesamiae from Mombasa and K itui regions. The difference in expression of CrV1 gene in B. fusca and S. calamistis suggests the involvement of the CrV1 gene in immune suppression and sheds light on the refractoriness of B. fusca towards C. sesamiae biotypes. Genetic Structure and Analysis of Striga gesnerioides (Witchweed) from Senegal Charlotte Tonessia1,4*, W. Moctar2*, B. Gowda3, A. Severin4, J. F. Rami1, N. Cisse1,2, M. Timko3 1 CERAAS (Centre d’Etude Re´gionale pour l’Ame´lioration de l’Adaptation a` la Se´cheresse) BP 3320 Thie`s, Escale, Se´ne´gal; 2ISRA/CNRA, BP 53, Bambey Se´ne´gal; 3Department of Biology, University of Virginia, Charlottesville, VA 22904, USA; 4Plant Physiology Laboratory, Universite´ de Cocody, Abidjan, Coˆte d’Ivoire * Corresponding author. Email: [email protected] The parasitic angiosperm, Striga gesnerioides is a root parasite of wild and cultivated legumes, among which cowpea (Vigna unguiculata) is suitable host. In West Africa, it causes serious yield losses up to 100% in some instances. Based upon the differential resistance response of various cultivars and breeding lines, five distinct races of S. gesnerioides parasitic on cowpea have been proposed to exist in west and central Africa. But the race of the parasite present in Senegal is not well known. Understanding the genetic structure and host-parasite interaction of S. gesnerioides from Senegal is important because it allows identification of races or biotypes thus improving chances of breeding success. Using amplified fragment length polymorphism (AFLP) profile analysis and statistical clustering methods, we have examined the genetic variability and phenotpicic relationships of 215 individual’s plants representing 12 different populations of S. gesnerioides collected from various areas throughout their suspected distribution range in Senegal. Our results showed that most genetic diversity resided among populations (91%) which are consistent with the breeding system of S. gesnerioides. Based on its genetic profile and host preference/resistance responses more race of the parasite seems to be present in Senegal. The most popular race of striga from Senegal and the race present in Mali are closely related. Analysis of Plasmodium falciparum var Gene Repertoire Expressed in Childrenwith Severe Malaria from Tanzania Joseph Paschal Mugasa1*, M. Rottman2, W. Qi2, H. Mshinda1, H.P. Beck1 1 Ifakara Health and Research Development Center, Ifakara, Tanzania; 2 Swiss Tropical Institute, Basel, Switzerland * Corresponding author. Email: [email protected] The var gene family of Plasmodium falciparum encodes for the variant surface antigen PfEMP1 (P. falciparum erythrocyte membrane protein 1). PfEMP1 is considered an important pathogenicity factor in P. falciparum infection by mediating cytoadherence to host cell intermediate groups (B/A and B/C) on the basis of their genomic location and upstream sequence similarities in coding and non-coding upstream regions. Severe malaria may be caused by parasites expressing restricted subset of PfEMP1 variants that confer optimal sequestration in immunologically naive individuals. Studies on var gene expression are difficult due to the broad variability of var genes in field isolates. A case control study was set in Ifakara, Tanzania, children under age of five years with severe malaria were recruited using WHO 2000 guideline of case definition, and children with

asymptomatic malaria were enrolled in a study as a control group. Expression patterns of different fragments of var genes sequences were compared. We found that P. falciparum isolates from children with severe malaria predominantly transcribe var genes with DBL1like domains that are characteristic of Group A or B/A var genes with reduced cysteine residues. In contrast, isolates from children with asymptomatic malaria predominantly transcribe var genes with DBL1-like domains that are characteristic of the B- and C-related var gene groups. Initial phylogenic analysis showed that dominant expressed var genes from severe malaria isolates cluster together, if compared with dominant expressed var genes from asymptomatic isolates. These preliminary results support the hypothesis that var genes with DBL1-like domains (Group A or B/A) may be implicated in the pathogenesis of severe malaria. K-Means Clustering and the Analysis of Malaria Microarray Data (MMD) Victor Chukwudi Osamor1*, E. Adebiyi1, S. Doumbia2 1 Department of Computer and Information Sciences (Bioinformatics Unit), College of Science and Technology, Covenant University, P.M.B 1023, Ota, Ogun State, Nigeria; 2Malaria Research Training Center (MRTC), Bamako, Mali *Corresponding author. Email: [email protected] On regular basis, enormous amounts of genomic data are generated by researchers. Forming a large part of these huge raw data resources is the malaria microarray data arising from the study of gene expression in both Plasmodium infected mosquito and human beings to understand their biology and evolve new drug remedies. Malaria, caused by Plasmodium falciparum, is lethal and responsible for major losses and death in sub-Saharan Africa. The purpose of microarray is to measure the concentration of mRNA for thousands of genes at given time points. Since data analysis have profound influence on interpretation of the final results, basic understanding of the underlying model surrounding computational tools is required for optimal experimental design and data analysis by biology researchers who are target users of such tools. k-means is one of the popular and simple partition computational models for clustering microarray data but has the problem of requesting for number of clusters from researchers who may not have idea about the structure of the data to be clustered. In this work, we analyse the operational iteration mechanism of kmeans by what we call Step-by-step k-means walk using synthetic gene expression data. However, we also examine the results of analysis of k-means and other clustering techniques on malaria microarray data (MMD) with the objective of revealing their strength and weakness on existing experimental works. We discover open problems that are both computational and experimental based, with K-means clustering having further problems of sensitivity to initialisation and also often find solutions that are local optimum and apparently far from global optimum. There is also the need for a new microarray experiment which ultimately raises some design issues. Our new simplified K-means algorithm for clustering (for example) gene expression data arising from malaria microarray experiment will be capable of fitting adequate number of genes clusters required for any given microarray input data set automatically without guessing k (number of clusters) input. Bioinformatics Analysis of Dynein Light Chains of Plasmodium falciparum Elijah K. Githui1*, E. De Villiers2, A.G. McArthur3

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386 1

Laboratory of Molecular Genetics, National Museums of Kenya, Kenya; 2Bioinformatics Facility, ILRI, Kenya; 3Marine Biological Laboratories, Woodshole, MA, USA * Corresponding author. Email: [email protected] Plasmodium species belong to the phylum apicomplexa. Plasmodium, Toxoplasma and Cryptosporidium are parasites of considerable medical importance while Theireria and Eimeria are animal pathogens with extensive impact on food production. Plasmodium falciparum is particularly important because it causes malaria that has 300–500 million clinical infections, mainly in sub-Saharan Africa and results in more than 1 million deaths each year. The malaria parasite actively invades the host cell in which they propagate and several proteins associated with the apical organelles have been implicated to be crucial in the invasion process. The biogenesis of the apical organelles is not well understood but several studies indicate that microtubulebased vesicular transport is involved. In P. falciparum, the microtubule destabilising drugs, colchicine, dinitoanilines and taxotere considerably reduce the number of new rings in culture or completely tachyzoites treated with oryzin or colchicine loses bulk of the subpellicular microtubules and the ability to re-invade host cells. At a higher drug concentration, the nascent spindle and the microtubules are disrupted and the parasites are incapable of nuclear division. Immunofluorescence studies indicate the presence of microtubule based proteins at the apical organelle and have suggested that cytoplasmic dyneins maybe involved in the biogenesis of this organelle, and also in the replication of these parasites. Other studies have shown the presence homologues of vesicle mediated transport (COP, Sec23/31, N SF) inside the parasite and in the trans-cellular transport in P. falciparum infected erythrocytes but their precise role is not investigated. Dynein is a multi-subunit complex composed of heavy chains whose C-terminus form a globular head containing multiple ATP-hydrolysis sites and a small stalk that binds to microtubules while the N-terminus is a flexible stem that interacts with intermediate chains and light chains to form the cargo binding domain of the enzyme. In this study, we analysed the light chains of P. falciparum because they provide adaptor surface to the cargoes and, therefore, there is likelihood for variations. The cytoplasmic dynein light chains consist of three different families: TcTex1/2, LC8 and LC7/roadblock. TcTex1/2 are differentially expressed in various tissues and have been implicated in the attachment of cargoes to dynein motor. Vertebrates TcTex1 binds directly to rhodosin and disruption of the protein leads to retinitis pigmentosa. The LC8 protein family is highly conserved and is found in many multimeric enzymes including nitric oxide synthase and myosin V. Complete loss of LC8 function is embryonic lethal. The LC7/roadblock family has been shown to bind directly to the intermediate chains of dynein complex. The data presented demonstrate that the P. falciparun dynein light chains sequence and functional domains show high sequence homology within the apicomplexa and that only the dynein LC 8 group2 (97 a.a.) has a high homology (e -40) to human while TcTex1 and dynein LC7 group1 (93 a.a.) have low homology (e -08, e -05, respectively) to the human homologues. This suggests that the sequence/domain differences in these proteins can be targeted for vaccine or drug design. Functional Studies of a Recombinant Anti-TcP2 Antibody in Trypanosoma brucei Benson Nyambega1,4*, C. Smulski1, M. Juri Ayub2, L. Simonetti1, V. Owino4, J. Hoebeke3, D. Masiga4, M. J. Levin1

379

1 Laboratorio de Biologı´a Molecular de la Enfermedad de Chagas (LaBMECh), Instituto de Investigaciones en Ingenierı´a Gene´tica y Biologı´a Molecular (INGEBI), National Research Council (CONICET), Buenos Aires, Argentina; 2Fundacio´n Instituto Leloir, CONICET, Buenos Aires, Argentina; 3Institut Cochin, De´partement Maladies Infectieuses, INSERM U567, Paris F-75014, France; 4International Centre of Insect Physiology and Ecology (ICIPE), P.O. Box 30772–00100, Nairobi, Kenya

* Corresponding author. Email: [email protected] The trypanosome ribosomal P proteins are located on the stalk of the ribosomal large subunit they where play a critical role during the elongation step of protein synthesis. The single chain recombinant antibody (scFv) C5 recognises the conserved C-terminal end of the Trypanosoma cruzi ribosomal P proteins although it possesses very low affinity for the corresponding mammalian epitope. We show that scFvC5 is able to specifically block the protein synthesis in vivo in Trypanosoma brucei, the causative agent for sleeping sickness in humans and nagana in cattle. In transgenic parasites, the mRNA corresponding to scFvC5 could be amplified by R T-PCR and the protein product was detectable by Western blot. Indirect immunofluorescence assay shows a cytoplasmic distribution similar to that obtained with purified scFvC5 on wild type parasites. Transfected parasites showed abnormal cell size, accompanied by growth arrest, consistent with a blockade of protein synthesis. Analogues to trypanosome scFv C5 antibody could be potential candidates as novel antiparasitic agents.

In Silico Design of Diagnostic Primers for the Identification of Phytopathogen of the Genus Xanthomonas Nelson Ndegwa, G. Lall, and D.-J. Kim* International Institute of Tropical Agriculture (IITA), P.O. Box 30709– 00100, Nairobi, Kenya * Corresponding author. Email: [email protected] Genome-scale comparisons of plant pathogen sequences are being made possible by the increasing availability of completely sequenced bacterial genomes. There is need to design pathogen diagnostic kits for the rapid and specific identification of newly emerging bacterial strains. These comparisons offer the opportunity of designing polymerase chain reaction (PCR) primers based on the variability of the intergenic regions for the amplification of products from target taxa and not from non-target taxa for the identification of pathogens at various taxonomic levels. In this study, an in silico bioinformatics approach is used to design primers from the Xanthomonas genomes available at the Comprehensive Microbial Resource website of The Institute of Genomic Research (http://www.cmr.tigr.org). Xanthomonas axonopodis pv. citri 306 was chosen as the reference species and the intergenic regions that are 400–800 bp in length were viewed to see genes that flank the 50 and 30 end of these regions, conserved and in synteny across the Xanthomonas campestris, X. axonopodis and X. oryzae species after performing a homology search. A multiple alignment of the intergenic regions flanked by these conserved genes was done and analysed using Amplicon software. Preliminary results support the logic that intergenic-targeted primers could be designed from the conserved region for specific identification of Xanthomonas bacteria at the genus, species and pathovar level.

380

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

The Identification, Isolation and Biochemical Characterisation of the Virulent Surface Protein (Pf-EMP1) from Placental Extracts in Pregnant Women and the Use of Synthetic Peptides in Malarial Vaccine Research Philbert Kasuku Onyango*, O.C. Onyango Egerton University, P.O. Box 536, Joro, Kenya * Corresponding author. Email: [email protected] The process whereby erythrocytes containing mature forms of P. falciparum adhere to micro vascular endothelial cells of vital organs and thus disappear from circulation is known as sequestration (cytoadherence). This process is mediated by a family of strain specific high molecular weight parasite derived proteins known as P. falciparum erythrocyte membrane protein 1. This protein is exported to the surface of infecting erythrocyte where it is anchored to a submembraneous accretion of parasite derived histidine-rich protein which becomes the points of attachment to the vascular endothelium. This adhesive protein is the only parasite protein unequivocally present on the outside of the erythrocyte. This project therefore, aims at identification, isolation and biochemical characterisation of this variant surface protein (PfEMP1) which mediates sequestration in placenta of infected women and is believed to be the main antigen determining the parasite population structure during chronic malarial infection of falciparum. The isolated protein will be characterised to determine the amino acid composition as well as the N-terminal amino acid sequence. Depending on the sequence obtained, chemically synthesised peptides will be used to study immunological reactions of the protein when induced as an immunogen. Monoclonal antibodies will be produced against the purified protein and since they are site specific reagents, they will be applied in the investigations of the protein to identify the immunogenic regions in the conformation of the protein. The synthesised peptides will act as an antigen for specific immunological purpose for example, vaccine production to intractable or difficult pathogens like Plasmodium. Screening for Anti-Plasmodium Activity Using Extracts from Extremophiles Found in Lake Bogoria, a Kenyan Soda Lake Patrick W. S. Okanya1*, S. Martin2, F.J. Mulaa1 1 Department of Biochemistry, University of Nairobi, P.O. Box 30197– 00100 Nairobi, Kenya; 2US Army – Walter Reed Project/KEMRI, KEMRI HQS, Nairobi, Kenya * Corresponding author. Email: [email protected] Recent studies have shown that microorganisms living in extreme environments are a potential source of medicinal compounds from their secondary metabolites. The Kenyan Rift Valley has several soda Lake and hot spring environments that are inhabited by extremophiles, whose medicinal value has not been investigated. This study therefore aimed at isolating and extracting secondary metabolites from extremophiles with bioactivity against Plasmodium falciparum parasites and eventual characterisation and molecular identification of these isolates. Water samples were collected from soda lakes and streaked on alkaline agar media. The bacterial colonies that grew were picked and replica plated on fresh agar plates. Forty-three pure colonies were isolated based on colony morphology and were designated as independent isolates. These colonies were incubated in liquid media for ten days at 45 8C to produce secondary metabolites. The metabolites were extracted by methanol and screened for anti-malarial activity. Five isolates with the highest activity against

P. falciparum were identified and classified using the 16S rDNA analysis. PCR amplification of the 16S rDNA produced a 1.5 kb fragment, which was sequenced. Phylogenetic analysis of the partial sequences showed that the isolates cluster together with Bacillus licheniformis, Bacillus sonorensis, Bacillus subtilis, and Pseudobacillus carolinae species. Identification and Characterisation of Deoxyribose Phosphate Aldolase (dpa) in Toxoplasma gondii Biscah Munyaka1*, Gabriela Colman2, Makoto Igarashi3, George Dautu3 1 Kenya Medical Research Institute (KEMRI), P.O. Box 54840–00200, Nairobi, Kenya; 2Servicio Nacional de Salud y Calidad Animal (SEN ACSA), 595-21-584-496, San Lorenzo, Paraguay; 3National Research Center for Protozoan Diseases, Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido 080-8555, Japan * Corresponding author. Email: [email protected] Toxoplasma gondii is an obligate intracellular parasite which has emerged as one of the most common opportunistic infections in HIV/AIDS patients. Parasites can escape immune attack, by differentiation into bradyzoites (cysts) which reside predominantly in the brain, and are able to persist for the life of the host. Toxoplasmosis in AIDS patients is considered to be as a result of reactivation of latent infection. In the expressed sequence tag database, deoxyribose phosphate aldolase (Dpa) was found to be specifically expressed in the T. gondii cyst (bradyzoites). We analysed the expression of T. gondii Dpa (TgDPA) in toxoplasma bradyzoites and tachyzoites by R T-PCR and for the first time we here by report that TgDPA is highly expressed in the Toxoplasma bradyzoites. The hypothesis of the present study was that dpa is involved in cyst formation and therefore the involvement of TgDPA in cyst formation is very crucial as an entry point in drug development against Toxoplasmosis based on cyst formation. A gene encoding TgDPA was identified in the T. gondii strain R H by comparative sequence analysis and gene cloning. The gene encoded a protein of 286 amino acids, having a predicted molecular weight of 31 kDa. The gene was cloned into pGEX-5X1 vector and the resulting recombinant plasmid pGEX-5X1/dpa was transformed into Escherichia coli DH5 and the recombinant protein was expressed then used for antibody production in mice and in rabbit. The antibody titers against the protein were analysed by western blotting. The results indicated the successful production of anti-Dpa antibody, which react with a 34 kDa protein. Immunofluorescent antibody test in the parasites showed that TgDPA was localised in bradyzoites cytoplasm. These findings suggest that dpa could be involved in cyst formation and has a potential to impact on public health initiatives. Trends in Plasmodium falciparum Genotype Mutations Associated with Antimalarial Drug Resistance in Kenya P. Liyala1, F. Eyase1, R. Achilla1, H. Akala1, J. Wangui1, J. Mwangi1, F. Osuna1, M. Wadegu1, Rosemary Nzunza1*, R. L. Coldren1, N. C. Waters2, S. Bedno1 1 United States Army Medical Research Unit-Kenya/Centre for Clinical Research, KEMRI, P.O. Box 29893-00202, Nairobi, Kenya; 2Division of Experimental Therapeutics, Walter Reed Army Institute of Research, Silver Spring, MD 20910, USA * Corresponding author. Email: [email protected] Plasmodium falciparum is responsible for the high mortality rate of malaria in sub-Sahara Africa. In an attempt to monitor this disease

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

and provide valuable information to the anti-malarial drug programme, we have surveyed malaria parasites from several geographically distinct areas in Kenya that represent varying degrees of drug resistance, drug usage patterns and levels of malaria transmission. Malaria specimens collected from Busia, Malindi, Mumias, and Kisumu were evaluated for drug susceptibility against a panel of 15 known anti-malarial drugs. DNA was isolated from these samples for the determination of specific mutations in the genes that have been correlated with drug resistance which include Pfdhfr, Pfdhps, Pfmdr1, and Pfcrt. We find a decrease in the number of mutations in Pfmdr1 and Pfcrt from parasites isolated from areas where chloroquine is rarely used to treat malaria. In contrast, parasites collected from areas that are believed to still treat malaria with chloroquine have substantial mutations in Pfmdr1 and Pfcrt. In vitro drug sensitivity assays demonstrate that a decrease in the number of mutations in Pfmdr1 and Pfcrt correlates with an increase in chloroquine susceptibility. In those areas of Kenya where Fansidar (Sulphadoxine– Pyrimethamine combination) has replaced chloroquine as the first line treatment for malaria, we find an increase in mutations in Pfdhps that corresponds to an increase in sulfadoxine resistance. We also did not observe any mutations at codons 164 for Pfdhfr and 581 and 613 for Pfdhps. The absence of these mutations is consistent with parasites isolated from East Africa during previous studies and reflects a lesser degree of drug resistance as that seen in Southeast Asia. Rate of Substitution, Date of Emergence and Speed of Dispersal of Rice Yellow Mottle Virus in Africa Fatogoma Sorho1,6*, A. Pinel-Galzi2, O. Traore3, M. S. Rakotomalala4, E. Sangu5, Z. Kanyeka5, E. Hebrard2, G. Konte3, Y. Sere1, D. Fargette2, S. Ake6 1 Africa Rice Center (WARDA), 01 BP 2031, Cotonou, Be´nin; 2IRD, BP 64501 F -34394 Montpellier Cedex 5, France; 3INER A, Laboratoire de virologie et de biotechnologie ve´ge´tale, 01 BP 476, Ouagadougou, Burkina-Faso; 4FOFIFA, BP289 Mahajanga 401, Madagascar; 5University of Dar es Salam, P.O. Box 35060, Tanzania; 6Laboratoire de Physiologie ve´ge´tale, UFR Biosciences, Universite´ Cocody-Abidjan, 22 BP 582 Abidjan 22, Coˆte d’Ivoire * Corresponding author. Email: [email protected] Rice yellow mottle virus sobemovirus (RYMV), causal agent of the rice yellow mottle disease transmitted by beetles, is at the origin of an emergent disease of rice in Africa. Isolates of each of the countries where the disease was observed were sequenced. The spatial structure of the diversity, the relations between genetic and geographical distance, as well as the phylogeny suggest an initial diversification of RYMV in East Africa, followed by a temporal spread from east to west of the African continent. The objective of this study is to assess the rate of substitution of RYMV, then to infer the dates of emergence and the speed of dispersal throughout Africa. In this study, we use BEAST, a cross-platform programme for Bayesian Markov Chain Monte Carlo (MCMC) analysis of molecular sequences. The rate of nucleotide substitutions (per site and per year) of the RYMV dates of divergence from temporally spaced sequence data were estimated by Bayesian inference from the sequences of a large collection of the virus (40 years). The temporally spaced data help to follow the accumulation of mutations over time and thus to estimate the mutation rate using coalescent-based population genetic inference. This rate of change, the first one for a plant virus, was compared with those of a set of animal and human viruses. It then served to calibrate the phylogenetic tree of RYMV. Dates of appearance of the virus in various regions

381

of Africa were determined under the hypothesis of relaxed molecular clock using Bayesian MCMC method implemented in BEAST. Finally, the speed of dispersal of RYMV throughout Africa was estimated. These estimations were compared with the epidemiological data on rice yellow mottle disease, with the history of rice in Africa and with the dispersal modes of the vector beetles to identify factors responsible of the emergence of the RYMV. Population Genetic Studies of Tsetse Flies (Diptera: Glossinidae): The Usefulness of Genomic Resources Johnson O. Ouma1*, G. Marquez 2, E.S. Krafsur2 Department of Biochemistry, Trypanosomiasis Research Centre, Kenya Agricultural Research Institute, P.O. Box 362–00902, Kikuyu, Kenya; 2Department of Entomology, Iowa State University, Ames, Iowa 50011, USA 1

* Corresponding author. Email: [email protected] Tsetse flies, Glossina spp. (Diptera: Glossinidae) are medically and economically important insects confined to sub-Saharan Africa. They are exclusively blood feeding and are the sole vectors of African trypanosomiases. Proper understanding of rates of gene flow and genetic differentiation among tsetse populations is key to the success of area-wide tsetse control using genetic methods. Ecological studies indicate that tsetse flies in the morsitans group have great capacity for dispersal and colonisation of suitable habitats. However, genetic data from nuclear and mitochondrial DNA markers seem to indicate otherwise. The apparent inconsistency between ecological and genetic estimates begs further research. We investigated the breeding structure of the tsetse fly, Glossina pallidipes at micro- and macrogeographic scales by analysing spatial and temporal variation at eight microsatellite loci. Our data suggest that there is little gene flow between populations separated by as few as 50 km leading to strong measures of genetic differentiation. These results show that G. pallidipes tend to remain close to their neighbourhoods. However, the number of simple sequence repeat (SSR) markers used in this study were very few and could have led to substantial inter- and intra-locus variance. Ongoing Glossina genome sequencing project offers an opportunity to mine the data for more SSR markers. The potential benefits of this approach are herein discussed. HIV/AIDS Susceptibility and Resistance Depends on HLA-A Alleles Diversity. Immunoinformatics Approach Muriira Geoffrey Karau1,*, R.M. Ngure1 P.K. Bundi2 1 Department of Biochemistry and Molecular Biology, P.O. Box 24 (20115), Egerton University, Kenya; 2Faculty of Health Sciences, Department of Human Anatomy, University of Nairobi, Kenya * Corresponding author. Email: [email protected] Genetic resistance and susceptibility to infectious diseases involves a complex array of immune-response and other genes with variants that impose subtle but significant consequences on gene expression or protein function. Numerous studies have identified a role for HLA genotype in AIDS outcomes, implicating many HLA alleles in various aspects of HIV disease. The complexity of the immune responses can be tackled by high thorough put computational methods. The current study examined how individual susceptibility and protections to HIV1/AIDS are affected by the diversity of HLA-A in the population. We eluted envelop glycoprotein of 10 HIV-1 isolates against 17 HLA-A alleles with an external major histocompatibility (MHC) binding prediction tool freely available in the site; http://mhcbin-

382

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

dingpredictions.immuneepitope.org. All the top 10 epitopes from each allele were calculated and ranked with the highest number of epitopes representing the resistant and lowest the susceptible. In the 17 alleles queried eight were susceptible to HIV-1/AIDS with number of epitopes less than 38, while nine were protective with epitopes number greater than 38. HLA-A *3002 allele had the highest number of epitopes suggesting that individuals with this HLA-A are likely to generate strong immune responses to HIV-1 envelop glycoprotein. Our data suggests that HIV-1 immune responses on the envelop glycoprotein depends on HLA-A allele. This approach is significant in determining epitopes that can be used in the design of the epitope-based vaccine and in epidemiological studies of resistance and susceptibility to HIV-1. Phylogenetics of HIV-1 in Heterosexual Cohort of Discordant Couples in Nairobi, Kenya Nzabanyi Fredrick Nindo1*, M. Matu2, C. Obuya1 1 Clinical Trials Laboratory, Department of Obstetrics and Gynaecology, Kenyatta National Hospital, Nairobi, Kenya; 2Centre for Microbiology Research (CMR), Kenya Medical Research Institute, Kenya * Corresponding author. Email: [email protected] Epidemiological and cohort approaches to HIV-1 transmission dynamics can now be complemented by phylogenetic data. HIV-1 transmissions in Africa occur between married adults who are discordant for their HIV-1 infection status. In Africa, all known HIV-1 genetic subtypes and groups including groups N and O are present. Whether the various groups, subtypes and recombinant forms of HIV-1 have biological differences with respect to transmissibility and the course of disease progression is not known and most studies have focused on human host immunological responses and not the evolutionary mechanisms of the HIV-1 virus. For this reason it is important to study the phylogenetic distribution of the HIV-1 genetic subtypes in Kenyan cohort of discordant couples with major focus on the HIV-1 seronegative partners who seroconvert during follow-up in prophylactic interventions and establish any epidemiological linkage. To date, 3000 couples attending voluntary counseling and testing centres (VCTs) in Nairobi, have been tested for HIV-1, of whom 8% were HIV-1 discordant, 9.2% were concordant HIV-1 positive and 82.8% were concordant HIV-1 negative. The HIV-1 discordant group will be enrolled and incidence and predictors of heterosexual transmission monitored at 3-month intervals for seroconversion of the seronegative partner. Gp120 and P24 regions will be amplified by PCR analysis of blood samples from both partners and sequenced. In the final set of experiments, evolutionary linkage to incidence will be assessed by phylogenetic tree analysis. It is expected that phylogenetic data derived from this study will shade important insights into HIV-1 transmission dynamics in discordant couples in the developing world. Rational Vaccine Design: The Challenges of Adenoviral Vectors C.O. Esimone1, Chukwuemeka Sylvester Nworu2*, M.U. Adikwu1, P.A. Akah2 1 Department of Pharmaceutics, Faculty of Pharmaceutical Sciences, University of Nigeria, Nsukka, 410001, Enugu State, Nigeria; 2 Department of Pharmacology and Toxicology, Faculty of Pharmaceutical Sciences, University of Nigeria, Nsukka, 410001, Enugu, Nigeria * Corresponding author. Email: [email protected] In our recent study, we developed HIV and adenoviral vectors as model probes for antiviral screening. These viral–vector-based assays were

found to be rapid, specific, reproducible and safe. Our present research focuses on the use of computational and predictive tools to circumvent the major challenges limiting the optimal use of adenoviral vectors in vaccine deliveries. Low pathogenicity, high infectivity and ability to accommodate large size transgene have made adenoviruses useful and popular candidate in vaccine delivery. The occurrence of circulatory adenoviral antibodies in human, induced by passive infection limits the vector titre and hence the delivered vaccine titre. The second challenge is the use of computational methods to circumvent the possibility of disseminated adenoviral diseases in replication-competent viruses and the problem of low vector titre with replication-defective vectors. The third challenge is to predictively construct adenoviral vectors with high ability to infect replicating and non-replicating cells. Bioinformatics Characterisation of Ligand-binding Hotspots in Plasmodium falciparum Daudi Jjingo1,2 Makerere University, Faculty of Computing and IT, P.O. Box 7062, Kampala, Uganda; 2Georgia Institute of Technology, School of Biology, (Jordan Lab), 310 Ferst Dr, Atlanta Georgia 30332, USA

1

Email: [email protected], [email protected] In this poster I will use an example of a Plasmodium falciparum surface protease (enzyme) to illustrate a putative methodology using existing bioinformatics software and tools, first to identify, and then characterise binding sites/hotspots on the protein surfaces of putative drug targets. These characterised binding sites can then be used for subsequent anti-malarial drug design. The WHO/TDR malaria strategic research project, through its product research and development, has identified several molecular targets in Plasmodium that could be exploited for the development of new drugs against malaria. These include the protease responsible for secondary processing of the P. falciparum merozoite surface protein (MSP-1). Though the TDR research and development has gone on to design and synthesise three possible inhibitors of this protease, there is still room for development of other inhibitors. Proper and accurate Drug (inhibitor) design is determined, among other things, by accurate knowledge of the target protein’s surface. I use sequence retrieval to obtain the protein sequence of the P. falciparum surface protease from the NCBI protein database. This is followed by sequence comparison (Blast) with the entire Protein Data Bank (PDB) (1, 2), to identify the closest protein (evalue must be >110) with known 3D structure. The identified protein is then taken through QsiteFinder, a binding site prediction method which uses an energy based method to predict binding spots on the protein’s surface. The predicted sites are then characterised by volume, coordinates and surrounding protein residues. This avails a reasonable catalogue of feasible binding hotspots on the P. falciparum surface protease, which could be used for drug (inhibitor) design. Exploring the Insect Acetylcholinesterase (AChE) Active Site Gorge: Toxicokinetic and AChE Sequences Analysis as Prospects to Molecular Design of Selective Insecticides J. Mutunga*, T. Anderson and J. Bloomquist Neurotoxicology Laboratory, Entomology Department, Virginia Polytechnic Institute and State University, Blacksburg, VA-24061, USA * Corresponding author. Email: [email protected] Most insects express two acetylcholinesterase (AChE) genes, ace-1 and ace-2. The ace-1 codes for AChE-1, which has a functional role in hydrolysis of acetylcholine during signal transduction in insect ner-

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

vous systems. Architecture of AChE active site gorge is characterised by functional motifs comprising a hydrophobic pocket, the H-bond network, the acyl pocket, and the peripheral site. Screening for differential interaction of bivalent ligands occupying both the catalytic and peripheral sites, with insect and vertebrate AChEs is important in the design of selective insecticides. Dimeric tacrines with varying methylene linkers (C2–C12) were used to probe B. germanica (CSMA strain), A. gambiae and human AChE active site gorge and data for bovine AChE was adapted from a study by Munoz-Ruiz et al. (2005). Dimeric tacrines were found to be more potent in vertebrate AChE, especially bovine, than in insects (Potency was ca. 16,000 and 82,500-fold higher for bovine AChE than for BgAChE and AgAChE, respectively). The potency of dimeric tacrines was more tether length dependent in vertebrate AChE than in insect AChEs. Insect AChE had unique C6 bump whereby the potency decreased seven-fold and twofold for BgAChE and AgAChE, respectively. Such a change occurs in a difference of a single carbon length indicating a significant interaction of the bivalents at C5 compared to C6 with the active site amino acid moieties. Further tests are on-going with other insects to confirm our hypothesis that this bump is unique in insects. We further performed a sequence analysis for the ace-1 sequence using the DeepViewSwiss-PdbViewer (http://expasy.org/spdbv/) to determine the 3D arrangement of catalytic signatures in the AChE active site gorge. We used ClustalW for alignment and homology analysis and the Treeview in the visualisation of the cladogram for the sequences. Possible applications of the ‘C6 bump’ and differential interaction of catalytic motifs of AChEs to the molecular design of highly potent, insect-specific anticholinesterases that are less toxic humans will be discussed. Maesanin: A Benzoquinone from Maesa lanceolata that Completely Inhibits Respiration in Bloodstream Trypanosoma brucei brucei Lucy Wanjiru Kariuki1,*, E.K. Nguu1, R.M. Njogu3, P.W. Kinyanjui1, J.O. Midiwo4, W.D. Bulimo1,2, J.K. Kiaira1 1 Department of Biochemistry, University of Nairobi P.O Box 30197, Nairobi, Kenya; 2US Army Medical Research Unit-Kenya, P.O. Box 29893-00202, Nairobi, Kenya; 3Basic Sciences Department, Botswana College of Agriculture, Private Bag 0027, Gaborone, Botswana; 4 Department of Chemistry, University of Nairobi P.O Box 30197, Nairobi, Kenya * Corresponding author. Email: [email protected] Maesanin a naturally occurring benzoquinone from plant Maesa lanceolata that inhibits growth of a wide range of microorganism. It has been shown to have antibacterial, antifungal antihelminth and more recently antitrypanosomal activity was established. This suggests that it could have some important pharmacological activity. To gain further insight into the mechanism of inhibition by maesanin, glucose catabolism in bloodstream and procyclic T. brucei brucei was studied. Respiration by the bloodstream forms of T. brucei brucei was studied at increasing concentrations of maesanin with glucose as the substrate. This was done by incubating about 5  107 trypanosomes at different concentration of measanin for 0–30 min. At 25 mM maesanin, all the parasites were immotile after 10 min. This was compared with motility when parasites were incubated with Salicyl Hydroxamic acid (SHAM), cyanide, antimycin and a combination of cyanide and SHAM. Maesanin unlike the other inhibitors showed complete inhibition. Respiration by the bloodstream forms of T. brucei

383

brucei incubated in phosphate saline glucose (5 mM, glucose) at pH 8.0 was studied at increasing concentrations of maesanin. It was observed that maesanin steadily inhibited the rates of oxygen consumption and the trypanosomes died rapidly within 5 min of incubation. At concentration of 2.5 mM maesanin per 108 trypanosomes, the O2 consumption was completely inhibited. Inhibition by (SHAM) and cyanide was about 80% and 90% under the same conditions respectively. However the parasites were not completely immotile as in the case of measanin. The rate of pyruvate production was also studied at increasing concentrations of measanin. It was observed that increasing concentration of measanin decreased the rate of pyruvate production by T. brucei brucei. The above results suggested that maesanin could be inhibiting a vital energy production step in these parasites. Sustaining Capacity Building and Implementing Bioinformatics at Institut Pasteur de Tunis Alia Benkahla*, I. Guizani, H. Mardassi, B. Bouhaouala, S. Ben Abderrazak, S. Ben Miled, L. Chargui, M. Chenik, S. Abdelhak, A. Ben Abdeladhim, K. Dellagi Institut Pasteur de Tunis, Tunisia * Corresponding author. Email: [email protected] Institut Pasteur de Tunis (IPT) is a Tunisian research institution founded in 1883. Its missions are to conduct Research and Training activities on infectious diseases, R/D on vaccines, and Public Health Laboratory activities (PHL). Research and training programmes are mainly oriented towards national health and/or economic problems such as rabies, leishmaniases, tuberculosis, viral hepatitis, bovine theileriosis, veterinary microbiology, genetic disorders and scorpion or snake envenomations. Molecular biology applications perfectly integrated in the research programmes of the Institute can be tracked back to the early 90s, providing a unique context for promoting bioinformatics and its applications at the national and regional level, in highly relevant research areas, particularly for Health. The diversity of the research activities, their complementarities, their unique combination within the health research system, the levels of expertise generated, and the North–South and South–South collaborations developed, contribute to this unique position. Bioinformatics is crucial and takes advantage of the last decade explosion of molecular biology and genomics data, to support and to be supported by the ‘‘wet research’’ laboratories. The implementation of bioinformatics at IPT was done in six steps, with the support and following the recommendations of the UNICEF/WHO/UNDP/World Bank Special Programme for Research and Training in Tropical Diseases (TDR), the Re´seau International des Instituts Pasteurs (RIIP), the Ministry of High Education, and the Ministry of Health: (i) PhD and Post-Doctoral training of a resource scientist having a mathematics background (September 1996–October 2004); (ii) continuous short term training of staff and students (from 1998 to date) through some of the high quality bioinformatics courses organised by TDR (n = 8), Institut Pasteur in Paris (n = 4), and ICGEB-Trieste (n = 3); (iii) a dynamics of bioinformatics research projects development has been initiated in 2001 through a project supported by TDR promoting institutional networking on leishmaniases and tuberculosis followed by other projects successfully funded by TDR and by Institut Franc¸ais de la Coope´ration or to be supported by the European Commission (SysCo project); (iv) upgrading the institute’s hardware (one PC per scientist, servers, backup system), software and network infrastructure (Local

384

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

network: Ethernet (10/100) and 1 Go for the backbone; internet connection: special line of 2 Mb/s) and dedicated buildings; (v) participating actively to the organisation of international introductory or advanced bioinformatics courses, and offers a national Master’s degree in bioinformatics; (vi) recruiting the trained bioinformaticist (December 2004), starting and equipping a bioinformatics group (March 2005). Four core functions have been assigned to the bioinformatics group: user support, bioinformatics support and establishment of an interface with ‘‘wet laboratories’’, in-house training, and R/D. As an added value to the development of its activities, the bioinformatics group have started collaborations with different bioinformatics centres (e.g.: the SANBI, South Africa, the Max-Planck Institute for Molecular Genetics, Germany, BIOBASE, Germany, IBDML, France) and is currently contributing to the development of adequate human resources. Keywords: Bioinformatics, Capacity building, Research, Training, TDR, RIIP, North–South collaborations, South–South collaboration, Tunisia. In Silico Prediction of Protein–Protein Interactions in MTB Infected Macrophages Ouissem Souiai1,2*, H. Essid1,3, I. Abdeljaoued3, M. Masigh4, J. de las Rivas5, H. Mardassi1, B. Jacq2, A. Benkahla1, C. Brun2 1 Institut Pasteur, Tunis, Tunisia; 2Institut de Biologie du De´veloppement de Marseille Luminy, France; 3Ecole Supe´rieure de la Statistique et Analyse de l’Information, Tunis, Tunisia; 4Institut Supe´rieur de l’Informatique, Tunis, Tunisia; 5Bioinformatics and Functional Genomics Research group, Salamanca, Spain * Corresponding author. Email: [email protected] Tuberculosis (TB) is responsible for about 3 million deaths each year. A third of the human population carries latent TB. The pathogen uses poorly understood complex strategies to subvert macrophage-killing activities. We will present here an in silico global integrative approach aiming at predicting protein–protein interactions in infected and non-infected macrophages (host–host, pathogen– pathogen and host–pathogen). To achieve this goal, the set of MTB protein–protein interactions available in the public domain and the large set of human protein–protein interactions available in the Agile Protein Interaction DataAnalyzer (APID: http://bioinfow.dep.usal.es/apid/index.htm) will be put into physiological context, using expression data from the Gene expression Omnibus database (GEO), literature mining and Gene Ontology annotations (GO). Then, using PPI network analysis methods based on graph theory PRODISTIN (Brun et al., 2003, Genome Biology; Baudot et al., 2006, Bioinformatics; http://crfb.univ-mrs.fr/webdistin/), the modular composition of each network will be determined and compared to each other. Differences in module composition between infected and noninfected macrophages are expected to point towards the biological processes which are preferentially targeted by the pathogen, activated in response to the infection and are responsible for the disease phenotype. Currently, a statistical test is being developed to rank the reliability of APID interactions in order to better predict the set of proteins that is effectively interacting in infected and non-infected macrophages. Keywords: MTB infected macrophages, Protein–protein interaction network analysis, Comparative genomics, Gene expression data, Literature mining

Building Functional Genomics Research Capacity on Insects Vector of Human Disease Vector in Africa Seydou Doumbia1*, M.B. Coulibaly1, S.F. Traore1, G. Dolo1, E. Adebiyi2 DMEVE/MRTC/Faculty of Medicine, University of Bamako, Mali; 2 Department of Computer and Information Sciences (Bioinformatics Unit), College of Science and Technology Covenant University, Ota, Ogun State, Nigeria 1

* Corresponding author. Email: [email protected] The genome sequences for many insects vector of human diseases are now available and promise the development of a set of new, powerful tools that can be used to develop innovative approaches to control these diseases. The African continent, which is the most severely affected by vector borne diseases, lacks adequate infrastructures and personal resources required for rational uses of genomic information. To fill this gap, the African Centre for Training in Functional Genomics of Insect vectors of Human Disease (AFROVECTGEN) was initiated by WHO/TDR and the Department of Medical Entomology and Vector Ecology (DMEVE) of the Malaria Research and Training Centre (MRTC) in Mali. The aim of the AFROVECTGEN programme is to train young scientists in functional genomics who will ultimately use genome sequence data for research on insect vectors of human diseases. Over the past 3 years, the African centre for training in insect disease vector functional genomics has contributed to the training of more than 40 junior scientists across the continent through a series of twoweek workshops sponsored by TDR/WHO. The training is focused on three main areas: Molecular Biology with emphasis on molecular entomology, Applied Bio-Informatics and Microarray Technology and Data Analysis. Research and training opportunities for carrier development in genomics research on insects vector of human diseases at the Malaria Research and Training Center will be presented. Keywords: Disease vectors, Bioinformatics, Functional genomics, Training, Africa

PhyML: Fast and Accurate Phylogeny Reconstruction by Maximum Likelihood S. Guindon, J. F. Dufayard, W. Hordijk, Vincent Lefort, O. Gascuel* Me´thodes et algorithmes pour la Bioinformatique, LIRMM, CNRS, University of Montpellier II, 161 rue Ada, 34392 Montpellier Cedex 5, France * Corresponding author. Email: [email protected] We will present and make a demo of PhyML, a software that implements a fast and accurate heuristic for estimating maximum likelihood phylogenies from DNA and protein sequences. This tool provides the user with a number of options (a large range of evolutionary models, estimation of various evolutionary parameters, bootstrapping, fast LRT of branch supports, several tree space search strategies, etc.), in order to perform comprehensive phylogenetic analyses on large data sets in reasonable computing time. This software is widely used and highly cited, and a new version (PhyML 3.0) will be available shortly. Users can download the binaries (Windows, Mac, Linux, and several other OS) or analyse their data via our web server. Papers, documentation, binaries, web server are available from http://atgc.lirmm.fr/phyml/.

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

References: [1] Guindon S, Gascuel O. PhyML–A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol. 2003;52(5), 696–704. [2] Guindon S, Le Thiec F, Duroux P, Gascuel O. PhyML Online: A web server for fast maximum likelihood-based phylogenetic inference. Nucleic Acid Res 2005;33, 557–559. [3] Hordijk W, Gascuel O. Improving the efficiency of SPR moves in phylogenetic tree search methods based on maximum likelihood. Bioinformatics 2005;21(24), 4338–4347. [4] Anisimova M, Gascuel O. Approximate likelihood-ratio test for branches: A fast, accurate, and powerful alternative (5). Syst Biol 2006;55(4), 539–552.

Functional Genomic and Structural Bioinformatics for Rational Drug Design Against Malaria in a Developing Country, Mali Modibo Coulibaly1*, B. Maigret2, S. Doumbia1, M.M. Devignes2, O. K. Doumbo2 1 University of Bamako, Faculty of Medicine, Pharmacy and Odontostomatology Malaria Research and Training Center, Mali; 2University of Henry Poincare Nancy, Faculty of Sciences and Techniques, Loria, France * Corresponding author. Email: [email protected] Despite major attempts over the past century to control malaria, this infection is still a life threatening disease affecting half a billion humans in underdeveloped and developing countries. Its global heartland is Africa, with an appalling death toll of 1–2 million people every year. Drug and vaccine research in malaria has a high priority. However, identification of suitable enzyme or antigens as candidates against malaria parasites development in human or in its vector mosquito has been on molecular biology mechanism of malaria life cycle. Genomic and post genomic studies has made the search of new drugs and vaccines easier. Urgent collaboration between endemic countries, and academic research institutions in developed country, i.e. required to address new and inexpensive drug discovery methods. We describe here a partnership development and implementation between the University of Bamako, Mali and the University of Henry Poincare of Nancy I, France. This collaboration aims to strengthen Malian capacity on computational biology and to create a platform and network between North and South for post genomic data integration and interpretation. Share data and allow research institutions to focus their clinical trials on the most promising potential targets. We aim to perform large scale functional genomic study to understand molecular mechanism that govern the Plasmodium life cycle. Genomic and proteomic data provided by functional genomic study can be integrated and interpreted by using computational power full tool and could allow us to identify suitable metabolic pathway targets. Structural bioinformatic tools can then be used to determine and predict protein structures and their molecular docking with drug candidate structures for the specific inhibitor research. We hope this collaboration will first strengthen our institute by allowing transfer of functional genomic and structural bioinformatics technology. In the long term we hope to develop new innovative drugs or vaccines that can inhibit parasite life cycle development in the host and vector.

385

In Silico Screening for Inhibitors of CLPQY Protease of Plasmodium falciparum Chijioke Uzochukwu Ikemefuna1*, Gupta Dinesh2 Department of Pharmaceutical and Medicinal Chemistry, Faculty of Pharmaceutical Sciences, Nnamdi Azikwe University, Awka, Anambra State, Nigeria; 2International Centre for Genetic Engineering and Biotechnology, New Delhi, India 1

* Corresponding author. Email: [email protected] The 3D molecular structure of ClpQy will be obtained from www.pdb.org. The physicochemical parameters of ClpQy (such as molecular weight, isoelectric point, etc.) will be determined using the SwissProt online machine. Ligand structures shall be obtained from the Cambridge Structural Database1, and/or Chemical Abstracts Online1, and/or LOPAC-12801, and/or TRANSFAC 3.51, etc. Rational molecular modification of known proteases inhibitors shall be done to afford other compounds for computational screening. Phenyl and heterocyclic structures will be selected using QUEST1. The generated structures will be screened against the ClpQy protein structure with DOCK3.51 in the contact score mode and structures with scores above acceptable value will be selected for further scoring based on steric fit within the active site of the ClpQy protease. The selectivity for the parasite enzyme over the human enzyme will be determined by substracting the contact score for parasite enzyme from the contact score for the human enzyme. Selected compounds will be visually screened for protein-ligand, electrostatic, H-bonding and hydrophobic interactions using MACROMODEL1. Chemical interactions will be assessed within DOCK 3.51 using the force field scoring. Successful compounds will be subjected to biological screening and in vitro assays. Lead protease inhibitors specific against Plasmodium falciparum ClpQy protease at a reasonable nanomolar range will be identified. In Silico Knock-out Screening of the Metabolic Network of Plasmodium falciparum to Yield Potential Drug Targets Segun Fatumo1,2,3, E. Adebiyi1*, R. Ko¨nig3 1 Computer and Information Science Department, Covenant University, PMB 1023, Ota, Ogun State, Nigeria; 2Cologne University Bioinformatics Center, University of Cologne, Germany; 3Bioinformatics and Functional Genome Analysis, Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Germany * Corresponding author. Email: [email protected] Plasmodium falciparum is responsible for about 90% of malaria deaths. Therefore, there is a need to discover new drug targets since the parasite has become resistant to existing malaria drugs. We have established a simple and efficient tool which analyses the metabolic pathways and as well identifies essential reactions/enzymes as possible drug targets in the metabolic network of P. falciparum. We investigated the essentiality of a reaction in the metabolic network of P. falciparum by deleting such a reaction in silico, 30% of the neighbouring compounds of the investigated reaction were chosen as products. We identified 203 essential reactions, our algorithm yield about 40% consensus choke-points as presented elsewhere. Consequently, we did not take essential reactions/enzymes as drug targets into account which have similarity to any human enzyme. Furthermore, our enzymatic reaction data were grouped into six clusters of coexpression of their corresponding genes during the red blood cell life cycle. We effectively identified 19 possible drug targets to be validated experimentally. Our

386

CONFERENCE ABSTRACTS / Infection, Genetics and Evolution 9 (2009) 368–386

aim is to combine inhibitors in the same cluster group of these possible drug targets additionally improving malaria therapy. Keywords: Knock-out, Essential, Plasmodium, Flux balance analysis DNA Array Based Differentiation and Identification of Leishmania Species, Encountered in Tunisia and the Mediterranean Region, Using Targets Identified In Silico Y. Saadi, A.S. Chakroun, M. Ben Fadhel, R. Lahmadi, A. Fathallah-Mili, S. Ben Abderrazak and Ikram Guizani* Laboratoire d’Epide´miologie et d’Ecologie Parasitaire: Applied biotechnologies and genomics to parasitic diseases research programme, Institut Pasteur de Tunis, Tunisia * Corresponding author. Email: [email protected] Leishmaniases constitute major public health problems in many countries over four continents, and are particularly endemic in Africa and the Mediterranean region. In Tunisia, the epidemiological situation is complex with proven existence of at least three Leishmania species: L. infantum, L. major and L. tropica, which are responsible for cutaneous leishmaniasis. Visceral leishmaniasis is caused by L. infantum. Simple, reliable, sensitive and rapid diagnostic tools still remain a research priority. The aim of this study is to develop diagnostic PCR based assays for the Leishmania species identification and discrimination using targets identified upon the comparative analysis of L. major and L. infantum genomes. We have exploited the genome project databases, available in the public domains: the L. major genome is completed while the L. infantum is still ongoing. We developed perl

scripts to Blast sequences from L. infantum against L. major and inversely. The sequences that were identified this way were then blasted on other databases like Genebank and EMBL. Following this selection process, two sequences were identified for L. major and four for L. infantum. The sequences were analysed in silico, they were assigned to six different Leishmania chromosomes (Chr: 8, 21, 23, 30, 32 and 33), and they were annotated using the ‘‘Leishmania bioinformatics toolbox’. Twenty-four primers were designed using the software ‘‘primer 3’’. Different primer pairs were considered for each sequence. Their specificity was checked in silico and was assessed experimentally. A total of 18 primer pairs were tested on three DNAs representative of the species L. major, L. infantum and L. tropica. Interestingly, specificity of four sequences was confirmed towards L. infantum (two) or L. major (two). For the remaining two, results are concordant with the latest genome releases. L. tropica DNA was amplified with all primer pairs tested with the exception of two primer pairs corresponding to one L. infantum marker. Three physically independent pairs were selected for the development of a multiplex PCR assay that allows the generation of three consistent and well differentiating amplification profiles, specific for the three Leishmania species encountered in Tunisia, L. major, L. tropica and L. infantum. A DNA array was further developed which allowed a drastic increase in sensitivity of detection and a significant reduction in results delivery time. Work is in progress to evaluate the potential of these tools for molecular diagnosis of Leishmaniasis in Tunisia. This work received financial support from MRSTDC-Tunisia (contrat programme 2004– 2008), from the RAB & GH programme of WHO/EMRO-COMSTECH and from AIEA Technical cooperation programme (TUN/06/12).